Final Report - Zixuan HE

6101SBC:
Engineering Project
Final report
Title: An ADRC Controlled

Standalone DC Microgrid
with Wind-PV Generation
and LIB Storage System
Name: Zixuan HE
LJMU number: 936370
Advisor: Dr. Karam Eliker

6101SBC: An ADRC Controlled Standalone DC Microgrid with
Wind-PV Generation and LIB Storage System
Abstract- Faced with rapid population growth and increased personal income, recent years have
witnessed a skyrocketing energy demand. However, the traditional energy structure based on fossil fuels
incurred global warming and climate aggravation. The economic development of emerging countries also
results in the fluctuation of fossil fuel prices. Hence, it is urgent to limit the dependency on fossil fuels and
accelerate the trend towards renewable energy sources(RESs). Distributed energy system (DES), as a new
energy supply concept, is playing a vital role at the background of carbon neutrality and energy
sustainability. This article proposes an ADRC controlled DC microgrid model based on wind-PV
generation system paired with a lithium-ion battery storage system to make it operate in standalone(off-
grid) mode, which is applicable to remote and rural places where cables are hard to reach.
Keywords: Hybrid renewable energy system, DC microgrids, MPPT algorithms, Energy Storage,
DC/DC Converters, Energy management
Zixuan HE, 936370

Table of Contents
List of Figures
List of Tables
Nomenclature
Chapter 1 Introduction ...............................................................................................................................1
1.1 Research background ............................................................................................................1
1.2 Problem statement .................................................................................................................2
1.3 Research objective .................................................................................................................3
1.4 Research contribution ...........................................................................................................4
1.5 Thesis structure .....................................................................................................................4
Chapter 2 Literature Review......................................................................................................................6
2.1 Introduction...........................................................................................................................6
2.2 Stand-alone MG system .......................................................................................................6
2.3 Energy management strategy ..............................................................................................8
2.3.1 Centralized control ................................................................................................................... 8
2.3.2 Decentralized control ............................................................................................................... 9
2.3.3 Hierarchical control ............................................................................................................... 10
2.3.4 A comparison of EM Strategy............................................................................................... 11
2.4 Droop control for DC/DC converters ................................................................................ 11
2.4.1 Droop control ........................................................................................................................................ 12
2.4.2 Trade-offs ............................................................................................................................................... 14
2.5 Existing control techniques for standalone DC MGs .......................................................15
2.5.1 H∞-based control methods .............................................................................................................. 15
2.5.2 State-space control methods ............................................................................................................ 16
2.5.3 Sliding mode control-based control methods ............................................................................. 16
2.5.4 Backstepping-based control methods ........................................................................................... 16
2.6 MPPT controller for wind-solar system ............................................................................17
2.6.1 MPPT algorithm for Wind Unit ..................................................................................................... 17
2.6.2 MPPT algorithm for PV Unit .......................................................................................................... 18
2.7 Summary ..............................................................................................................................20
Chapter 3 Mathematical Modelling of DC MGs Components ..............................................................21
3.1 Introduction..........................................................................................................................21
3.2 DC/DC converter modelling ...............................................................................................21
3.2.1 Buck converter ..................................................................................................................................... 21
3.2.2 Boost converter .................................................................................................................................... 23
3.3 WT system modelling ..........................................................................................................24
3.4 PV system modelling ...........................................................................................................26
3.5 ESS module modelling ........................................................................................................28
3.6 Summary ..............................................................................................................................30
Zixuan HE, 936370
Chapter 4 MPPT algorithm......................................................................................................................31
4.1 Introduction..........................................................................................................................31
4.2 MPPT algorithm for WT system ........................................................................................31
4.2.1 P&O algorithm for wind system .................................................................................................... 31
4.2.2 INC algorithm for wind system ...................................................................................................... 33
4.3 MPPT controllers for PV system........................................................................................34
4.3.1 P&O algorithm for PV system ........................................................................................................ 35
4.3.2 INC algorithm for PV system .......................................................................................................... 36
4.4 Summary ..............................................................................................................................38
Chapter 5 Control Techniques for ESS Module .....................................................................................39
5.1 Introduction..........................................................................................................................39
5.2 ADRC scheme for ESS module ..........................................................................................39
5.3 EM strategy formulation.....................................................................................................41
5.4 Summary ..............................................................................................................................43
Chapter 6 System Simulation and Analysis ............................................................................................45
6.1 Introduction..........................................................................................................................45
6.2 WT system simulation .........................................................................................................45
6.3 PV system simulation ..........................................................................................................47
6.4 ESS module simulation .......................................................................................................50
6.5 DC voltage simulation .........................................................................................................52
6.6 Summary ..............................................................................................................................54
Chapter 7 Conclusion and Outlook .........................................................................................................55
7.1 Conclusion ............................................................................................................................55
7.2 Further research objective ..................................................................................................55
Reference .....................................................................................................................................................56
Appendix .....................................................................................................................................................65
Acknowledgement ......................................................................................................................................66
Zixuan HE, 936370

List of Figures
Figure No. Figure Title Page No.

1-1 The considered DC microgrid. 3
1-2 Thesis structure of this article. 5
2-1 Standalone MGs with (a) common DC bus; (b) common AC bus. 6
2-2 The hierarchical structure of a MG control system. 8
2-3 Layout of centralized control approaches. 8
2-4 Layout of decentralized control approaches. 9
2-5 Layout of hierarchical control approaches. 10
2-6 Operating principles of DC/DC converters. 12
2-7 The control scheme for Voltage-Mode approach. 12
2-8 The control scheme for Current-Mode approach. 13
2-9 Inner loops of DC/DC converters. 13
2-10 Optimization of secondary control on DC voltage in DC MGs. 14
2-11 MPPT for WT system based on speed regions. 17
2-12 I-V and P-V characteristics of a PV module. 19
3-1 Working condition of a buck converter. 22
3-2 Working condition of a boost converter. 23-24
3-3 Model of WT system. 24
3-4 Dynamic model of the wind generator 26
3-5 Model of PV system. 26
3-6 Equivalent electrical circuit of a solar cell. 27
3-7 Model of ESS module. 28
3-8 Working condition of LIB in (a) Charging condition; (b) Discharging 29
condition.
4-1 MPPT principles based on variable wind speeds. 31
4-2 P&O method with (a) fixed step; (b) adaptive step. 32
4-3 Flowchart of improved P&O MPPT algorithm. 32
4-4 Operating principle of INC method in WT system. 33
4-5 Flowchart of INC MPPT algorithm with adaptive step in WT system. 34
4-6 I-V characteristics of a PV module under (a) Changing irradiance at 34
25℃; Changing temperature at 1000𝑊/𝑚2 .
4-7 P-V characteristics of a PV module under (a) Changing irradiance at 35
25℃; Changing temperature at 1000𝑊/𝑚2 .
4-8 (a) Flow chart of P&O MPPT algorithm; (b) Path of MPP under 35
changing irradiance.
4-9 Operating of principles INC method in PV system. 37
4-10 Flowchart of INC MPPT algorithm in PV system. 38
5-1 ADRC control structure block diagram 39
5-2 ADRC based DC-voltage control structure block diagram. 40
5-3 PI-based DC-voltage control structure block diagram. 41
Zixuan HE, 936370

5-4 Working condition of ESS module at discharging mode. 43
5-5 Working condition of ESS module at charging mode. 43
5-6 The EM structure of the DC microgrid. 44
6-1 (a) Simulink model of wind generator; (b) Simulation curve of wind 45
generator.
6-2 Simulink model of boost converter in WT system. 46
6-3 P&O MPPT controller submodule in WT system. 46
6-4 Simulation curve of MPPT controller in WT system. 47
6-5 Comparison of P&O method with fixed step and adaptive step in WT 47
system.
6-6 Simulink model of PV panel. 48
6-7 P&O MPPT controller in PV system. 48
6-8 Simulink model of boost converter in PV System. 49
6-9 Simulation curve of MPPT controller in PV system. 49-50
6-10 Simulink model of bidirectional converter in ESS System. 51
6-11 Simulink model of control part in ESS module. 51
6-12 Simulation curve of LIB sector. 52
6-13 Simulink model of EM strategy. 52
6-14 Simulation curve of DC bus voltage. 53
6-15 Simulink model of proposed standalone DC microgrid. 54
List of Tables
Table No. Table Title Page No.

2-1 Challenges and solutions for standalone MGs. 7
2-2 A comparison of EM strategies. 11
2-3 Existing control techniques for standalone DC MGs. 15
2-4 Existing MPPT algorithms for wind unit. 18
2-5 Existing MPPT algorithms for solar unit. 19
4-1 Truth table for improved P&O algorithm. 36
5-1 Working condition of each DC MG components. 42
Zixuan HE, 936370

Nomenclature
Abbreviation Full Name

DES Distributed Energy System
RES Renewable Energy Sources
DG Distributed generation
PV Photovoltaic
WT Wind turbine
EM Energy Management
PMSG Permanent Magnet Synchronous Generator
CCM Controller Current Source
LIB Lithium-Ion Battery
VRLA Valve Regulated Lead Acid
MG Microgrid
ESS Energy Storage System
HESS Hybrid Energy Storage System
MPP Maximum Power Point
P&O Perturbation and Observation
INC Incremental Conductance
HRES Hybrid Renewable Energy System
ADRC Active Disturbance Rejection Control
IPC Indirect Power Controller
DPC Indirect Power Controller
SoC Socket of Charge
DoD Depth of Discharge
TD Tracking Differentiator
ESO Extended State Observer
NLSEF Non-linear State Error Feedback
UC Ultracapacitor
CPU Central Processing Unit
CC Central Controller
LC Local Controller
IoT Internet of Things
P2P Peer-to-Peer
LP Linear Programming
LMI Linear Matrix Inequalities
IBC Integral Backstepping Control
SMC Sliding Mode Control
DC Direct Current
AC Alternating Current
Zixuan HE, 936370

Chapter 1 Introduction
1.1 Research background
In most remote areas around the world, power supply is mainly supplied from diesel generation. Due to
the high cost and infeasibility of connecting local loads to utility grids, this type of distributed
generation(DG) remains prevalent(Jing et al., 2016). However, given global warming, anthropogenic
pollutant emissions, fuel price fluctuation, and carbon neutrality initiatives, continued utilization of fossil-
fuel based generation systems has become no longer reasonable(Zanganeh and Shafeen, 2007). In addition,
geographically isolated areas like islands and plateaus tend to have a great reserve of wind and solar
energy, and these renewable energy sources(RESs) are an excellent alternative for diesel. Hence, the
transition from traditional fossil fuel to RES has been an inevitable trend. The use of RESs in DG systems
is gaining popularity to cope with the deficit of electricity supply, particularly in areas where existing
power grids are difficult to reach(B. S., Yan and Rudy, 2013).
However, RESs have their defects. Photovoltaic(PV) systems rely on irradiance and temperature, wind
turbine(WT) system is limited to variable wind speed and air density, and hydropower is restricted by
altitude elevation and amount of precipitation(Bari et al., 2014). The unpredictable nature of RESs makes
the intermittency of electricity supply a challenging task to deal with. A microgrid(MG) is a small-scaled
regional power system that consists of local energy generation systems, energy storage systems(ESS), and
load demand, with an objective to ensure the stable power flow to meet local demand(Lin, Zhao and Wu,
2020). An MG system can be either DC or AC or a combination of both, but DC-based options are
becoming more mainstream mainly because DC MGs can run in standalone(off-grid) mode, at which
consistent power can be supplied to meet local demands autonomously without the need of the prohibitive
long-distance high-voltage transmission infrastructures(Jing et al., 2016). Hence, the integration of RESs
into DC MG systems has been a potential solution to cope with the deficit of electricity supply in remote
regions(Ullah et al., 2020).
Renewables-based DC microgrid is being opted for and studied universally because of its high energy
Zixuan HE, 936370

1
efficiency and widespread applications(Luna et al., 2017). Among those RESs, solar power and wind
power are the most preferred due to their easy availability and advanced Maximum Power Point
Tracking(MPPT) technologies(Hossain et al., 2019). As for storage medium, the valve-regulated lead-
acid(VRLA) batteries are widely utilized because of the low expense and easy manipulation(Jing et al.,
2016). Since RESs can complement each other, MG systems composed of two or more RESs could have a
positive effect on the storage capacity(Atia and Yamada, 2016). In the future, liquified hydrogen is
projected to be the ultimate energy storage medium due to its clean, pollution-free and high energy density
properties(Li, 2018).
In DC MG systems, the stability of the DC bus voltage is of utmost importance. Power converters are
the core part of hybrid renewable energy systems(HRESs) because they stabilize the voltage output during
power intermittency(Farhani, Barhoumi and Bacha, 2021). The power quality of HRESs heavily relies on
the stable operation of controllers for power converters. For instance, an MPPT controller for boost type
power converters is widely used in PV systems to keep the DC voltage stable(Padmanaban et al., 2018).
Among the mainstream power conversion devices, DC/DC converters are proved more efficient due to
their simple structures, easy manipulation, and high efficacy(Sivakumar et al., 2016)(Meraj et al., 2021).
In MG systems, MPPT algorithms are widely implemented on DC/DC converters to extract optimal output
power, and ESSs are required to store and release excessive energy produced by RESs. Perturbation and
observation(P&O) and incremental conductance(INC) algorithms are the most used MPPT controllers,
which have been long experimented and applied to commercial utilization(Mark, 2010).
1.2 Problem statement
In north-western China, where solar energy and wind energy is rich, there is a mismatch between
renewables distribution and energy demand. Most electricity generated by renewables is wasted so that
renewables-based DC MG is a solution to address the power redundancy and intermittency. Moreover, due
to the rapid development of lithium-ion batteries(LIB) technology these years, LIBs have the potential to
revolutionize the ESS in HRESs(Smith et al., 2017). Combined with context, the problems to be solved in
this research are proposed as:
Zixuan HE, 936370

2
(1) The power converter modelling for DC microgrid in standalone mode
(2) The integration of WT system, PV system, and LIB sector into DC/DC converter
(3) The programming of MPPT algorithm in the controllers for WT and PV systems, and the design of the
ADRC controller to control the LIB sector
(4) The formulation of EM strategy in ESS module to switch the DC microgrid to different modes
1.3 Research objective
Fig. 1-1. The considered DC microgrid.
The main goal of this article is to design a DC microgrid model based on wind-PV generation and LIB
storage systems. From the perspective of energy conversion, this MG model includes three parts: energy
generation, energy storage, and energy consumption(Lu and Fu, 2011). As shown in Fig. 1, the considered
DC microgrid mainly consists of PV system, WT system, LIB sector(ESS module), a DC bus, and a load.
The following are the specific experimental objectives to be done:
(1) Design a dynamic model of wind generator paired with PMSG in WT system
(2) Design a boost converter for WT and PV systems, and a bidirectional converter for the ESS module.
(3) Design an MPPT controller for WT and PV systems using P&O algorithm, and an ADRC controller
for the LIB sector to stabilize the current flow in the battery sector.
Zixuan HE, 936370

3
(4) Formulate EM strategy in the ESS module together with the ADRC controller to switch the DC MG
system to charging mode or discharging mode.
1.4 Research contribution
In the DC microgrid proposed in this article, more emphasis is put on the modelling of DC/DC
converters for wind energy system, solar energy system, and LIB energy storage system. And the main
achievements of this DC microgrid research are listed as followed:
(1) Bertz’s Law is verified based on the mathematical modelling of wind generators with PMSG.
(2) Operating principles of wind turbines, photovoltaic cells and Lithium-ion batteries are analyzed.
(3) Presented the overall structure of each DC microgrid component based on DC/DC converters
(4) The comparison of P&O and INC MPPT algorithms is carried out for the PV system, and the ADRC
scheme is compared with PI controller to regulate the DC voltage more intelligently.
(5) The simulation of WT system, PV system, and ESS module are demonstrated in Simulink, and the
output characteristics are analyzed to provide a theoretical and simulation basis for further study of
DC MG systems.
1.5 Thesis structure
This article is based on a research on a standalone(off-grid) DC microgrid model, including the working
principles of HRESs, power converter modelling of MG system, and control techniques for each section.
The contents covered in this article are organized as follows and depicted in Fig. 1-2:
In chapter 1, the background of this research is briefly introduced, and then the problems to be solved in
this research, the aims and main achievements of this research are listed in items. In chapter 2, a review is
done on the current status of research on MPPT and DC microgrid configuration at home and abroad. In
chapter 3, the mathematical modelling of DC microgrid components, including WT system, PV system
and LIB sector are presented based on DC/DC converters. Chapter 4 demonstrates the MPPT algorithm
for wind and solar system. Then chapter 5 presents the ADRC scheme used for battery control and
formulates the EM strategy in ESS module. Chapter 6 demonstrates the simulation and experimental
Zixuan HE, 936370

4
results based on Simulink models. Finally, chapter 7 concludes the article and proposes the research
forecast.
Fig. 1-2. Thesis structure of this article.
Zixuan HE, 936370

5
Chapter 2 Literature Review
2.1 Introduction
Due to the unpredictable nature of wind and solar energy, their higher penetration in standalone MG
systems can lead to high technical challenges, especially the lack of adequate storage capacity.
Fluctuations of voltage, current, and frequency are the main power quality issues for MG systems in
standalone mode(Badwawi, Abusara and Mallick, 2015). In this chapter, an introduction is first presented
for standalone MG systems, including system configuration, challenges and proposed solutions. Then a
review is conducted for EM strategy, control techniques for islanded DC MGs and MPPT algorithms.
2.2 Standalone MG system

The standalone/off-grid/autonomous MG systems are a smart solution for remote regions where power
transmission lines and distribution infrastructures are uneconomic to operate or hard to install due to
prohibitive cost and geographical limitations. As shown in Fig. 2-1, according to different converter
configurations, the standalone systems can be categorized into a common DC or AC bus, where all the
generated power is gathered and further dispatched according to the energy management(EM) strategy. In
such a hybrid MG system, wind energy can make up for the deficiency in solar energy over a certain
period to partially overcome the volatility of RESs. The integration of two complementary RESs can
minimize the sizing of storage elements and optimize the whole system to a more flexible and reliable one.
Fig. 2-1. Standalone MGs with (1) common DC bus; (2) common AC bus.
For standalone applications, the cost of ESS module accounts for the major economic issue. Since
Zixuan HE, 936370

6
batteries have a much shorter lifespan in contrast to a PV or WT, increasing PV panels and wind turbines
could be an obvious choice to minimize the storage capacity and ultimately reduce the overall cost(Yang,
Lu and Zhou, 2007). While for HRESs that require high reliability, too few batteries cannot meet the
storage requirement. Hence, battery storage and diesel generator backup systems become a popular
solution for RES redundancy during the year(Belfkira, Zhang and Barakat, 2011). However, if the ESS
module runs out of power, there is no source to import energy from. Moreover, the negative impact of
diesel engines on the environment has incurred worldwide controversy. Therefore, integrating RESs with
fuel cells is a promising alternative for standalone MGs, in which fuel cells acts as a storage medium to
stabilize fluctuations in power supply by storing the excess energy generated by DG systems and
delivering power to the load demand during the intermittent period(Nelson, Nehrir and Wang, 2006).
Various control techniques have been proposed that could be applied to achieve a techno-economical
standalone MG system.
Table. 2-1 Challenges and solutions for standalone MGs.
No. Challenges Solutions

1 High storage cost A combination of solar and wind energy minimizes storage capacity
and reduces the overall cost.
2 Insufficient usable The integration of battery storage systems and diesel generators into
energy HRESs can provide energy during an intermitted period.
3 Power intermittency/ Battery or fuel cell storage systems serve as major energy sources,
power quality and in some cases, diesel generators provide backup power.
4 Protection Protection schemes or devices are installed and upgraded, such as the
introduction of distributed generators.
5 Storage shortage Fuel cells and supercapacitors are integrated to upgrade the ESS.
6 Environmental and VRLA batteries or ultracapacitors are replaced by new fuel cells such
safety concerns as LIBs, leading to fewer contaminants and less total maintenance
towards batteries costs.
Zixuan HE, 936370

7
2.3 Energy management strategy
Fig. 2-2. The hierarchical structure of an MG control system.
Energy balance in HRESs requires a reliable control structure. As shown in Fig. 2-2, a pyramid-
structured control system could manage energy dispatch intelligently and improve the flexibility and
reliability of MG systems. As HRESs continuously grow, more scientific EM approaches are required for
an orderly integration within current electricity networks. Up to now, there are three mainstream EM
strategies: Centralized control, decentralized control, and hierarchal control.
2.3.1 Centralized control
Centralized control scheme uses a central controller(CC), which is characterized by a high-performance
CPU to manage the power dispatch of all entities of the system with communication networks, as shown
in Fig. 2-3. And the control decisions generated by CC are directly sent to local controllers(LCs) in each
entity to interact with the CC. As the core part, the CC collects data, ranging from RESs production to
power consumption to energy price to weather conditions, etc, so as for optimal and efficient energy
control(Warnier et al., 2017).
Fig. 2-3. The layout of centralized control approaches.
Zixuan HE, 936370

8
Moreover, using emerging communication and computation technologies such as IoT and Big-Data, the
CC could accomplish the supervision, collection, and analysis of data feedback. This allows all entities to
synchronize with the CC and ensures a flexible MG operation in both on-grid and island modes
(Elmouatamid et al., 2020). However, despite the easy implementation, CC based solutions have shown
their limitations, especially when dealing with large-scale hybrid systems.
2.3.2 Decentralized control
In decentralized control, each entity is operated autonomously by its own LC. As shown in Fig. 2-4,
each LC operates individually to manage RESs, ESSs, and loads rather than be instructed by a CC unit.
The control decisions are determined locally on a basis of local measurements of frequency, voltage
values, and other system characteristics which are shared among the other LCs via peer to peer(P2P)
communication networks(Yamashita, Vechiu and Gaubert, 2020).
Fig. 2-4. The layout of decentralized control approaches.
According to the availability of communication networks, decentralized control can be subclassified
into three modes: (i) Fully dependent control, in which LCs are interacted to share information while the
decisions are generated by CC; (ii) Partially independent control, in which control decisions are made by
LCs, and CC acts as a communicating medium to deliver information to each other; and (iii) Fully
independent, in which LCs make decisions themselves and directly interact with each other, being entirely
Zixuan HE, 936370

9
independent of the CC(Celik et al., 2017). However, although decentralized control structure increases the
flexibility of operation, it presents a lower response time mainly because of the incomplete information
during MG system installation(Espín‐Sarzosa, Palma-Behnke and Núñez‐Mata, 2020).
2.3.3 Hierarchical control
In fact, the hierarchical control scheme is a compromise between fully centralized and decentralized
control structures(Van et al., 2020). Since fully centralized approaches require extensive communication
and computation infrastructure, which is a difficult task to implement on a large scale. Meanwhile,
decentralized control requires a high level of coupling and coordination, which is hard to achieve without
the participation of CC(Molzahn et al., 2017). Hence, the hierarchical control structure is proposed to
separate EM into three levels: Primary control, secondary control, and tertiary control(Van et al., 2020).
Fig. 2-5. The layout of hierarchical control approaches.
As depicted in Fig. 2-5, the primary control level stabilizes the measurement values generated from
RESs to follow the requirements regulated by the standards(Elmouatamid et al., 2021). Also, the primary
control section detects the system status and switches the mode of MGs(Prabaharan et al., 2018). In
secondary control, the system characteristics(voltage, current, frequency, etc.) are restored after load
variations so that the power quality can be maintained and the synchronization between MGs and the
Zixuan HE, 936370

10
utility grid can be ensured (González-Romera et al., 2020). The tertiary control is in charge of controlling
the power flow, so as to ensure the capacitance and inductance operate in optimal performance in both
modes(Guerrero et al., 2013).
2.3.4 A comparison of EM strategy
Table 2-2. A comparison of EM strategies.
EM Strategy Control Logic Merits Limitations

Centralized A CC manages all Strong controllability and CC failure affects the
control entities by controlling real-time observability whole system operation
LCs with
communication Suitable for small size MG Large-scale
networks. systems communication and
computation infrastructure
Decentralized LCs tend to operate LCs provide autonomous Requires a high level of
control without a CC control capability coupling and coordination
Three types: Fully P2P network allows flexible High complexity of

dependent, partially operation and avoids single- implementation and
independent, and fully point failure require fast periodical
independent reconfiguration
Hierarchal Three control levels: Flexible regulation of system No energy is transferred if

Control Primary, secondary, characteristics within an information fault exists
and tertiary acceptable intervals in the upper layer
A combination of More suitable for DC MG More generators to control

centralized and systems voltage regulation and
decentralized control frequency
2.4 Droop control for DC/DC converters
DC/DC converters are extensively utilized in HRESs. As illustrated in Fig. 2-6, the DC/DC converter is
operated paired with the control technique, to regulate the load demand and power source simultaneously.
Zixuan HE, 936370

11
Furthermore, the control technique compares the feedback from the load with the desired reference value
to conduct optimal operation of the DC/DC converter.
Fig. 2-6. Operating logic of DC/DC converter.
2.4.1 Droop control
Droop control is normally applied for frequency control of generators and sometimes for voltage
control. It can also be used to share current in MG systems. Droop control can be divided into two modes:
voltage control mode and current control mode. In voltage control mode, the converter operates as a
controllable voltage source with voltage reference as the input. Alternatively, the converter in the current
control mode works as a controllable current source(Gao et al., 2019).
1) Voltage-Mode approach
Fig. 2-7. The control scheme for Voltage-Mode approach.
As shown in Fig. 2-7, the DC voltage reference is generated according to the output DC current(𝐼0 )
using the 𝑉 − 𝐼 droop characteristic:

∗ (2 − 1)
𝑉𝑑𝑐 = 𝑉𝑜 − 𝐼𝑑𝑐 𝑘
∗
𝑉𝑜 is the nominal bus voltage, 𝑘 is the droop gain, 𝐼𝑑𝑐 is the measured DC current, and 𝑉𝑑𝑐 is the
generated DC voltage reference.
Zixuan HE, 936370

12
2) Current-Mode approach
Fig. 2-8. The control scheme for Current-Mode approach.
As shown in Fig. 2-8, the current-mode approach uses the measured voltage to calculate the desired DC
current. The DC current reference is derived from the 𝐼 − 𝑉 droop characteristic based on the measured
DC voltage:
∗
𝑉𝑜 − 𝑉𝑑𝑐
𝐼𝑑𝑐 = (2 − 2)
𝑘
∗
𝑉𝑑𝑐 is the measured DC voltage, 𝐼𝑑𝑐 is the generated DC current reference.
3) Inner loop
The inner loop of the DC/DC converter differs according to different control modes. Based on different
input and reference variables, the inner loop generates different duty cycles as an output. As shown in Fig.
2-9, there are three basic types of inner loops: single voltage loop, single current loop, and a combination
of the first two named cascade loop.
(a) Single voltage loop. (b) Single current loop.
(c) Cascaded loop.
Fig. 2-9. Inner loops of DC/DC converters.
Zixuan HE, 936370

13
Fig. 2-9(a) shows a single voltage loop when the converter is working in voltage control mode. The
output of the voltage controller 𝐺𝑣 is the duty cycle. 𝑉𝑜_𝑟𝑒𝑓 is given by a 𝑉 − 𝐼 droop characteristic. Fig. 2-
9(b) shows the inner loop in the current control mode. The current reference(𝐼𝑟𝑒𝑓 ) is generated by the 𝐼 −
𝑉 droop characteristic to attain the duty cycle. And figure 2-9(c) shows another control mode, in which the
voltage controller 𝐺𝑣 provides the inductor current reference(𝐼𝐿∗ ) and the duty cycles are given by the
output of the current controller 𝐺𝐼 (Lakshmi and Hemamalini, 2017).
2.4.2 Trade-offs
As a basic control approach for achieving ideal power sharing, droop control enhances the modularity
and reliability of the system. However, because of the decentralized property, the impedance on the
distribution line has a negative impact on the control performance. Also, the cable resistance is sensitive to
climate changes, especially temperature variation. Moreover, the nominal voltage difference could result
in a degradation in load sharing. Thus, advanced droop control methods including inversed droop control,
non-linear droop control, dead-band droop control, and adaptive droop control are proposed and
developed(Kim, Choi and Cho, 2002; Chen et al., 2015; Xu, Sha and Liao, 2015; Prabhakaran, Goyal and
Agarwal, 2018).
Fig. 2-10. Optimization of secondary control on DC voltage in DC MGs.
Hence, the secondary control is proposed to minimize the deviation of voltage and improve the power
quality(Dragicevic et al., 2014). As shown in Fig. 2-10, when the primary control is utilized, the operating
point will deviate from 𝑉𝑜 (voltage at no load condition) to 𝑂𝑃1 under 𝐼𝑑𝑐1 and from 𝑉𝑜 to 𝑂𝑃2 under 𝐼𝑑𝑐2.
Zixuan HE, 936370

14
After integrating the secondary control, the operating point will shift from 𝑂𝑃1 to 𝑂𝑃1_𝑛𝑒𝑤 and 𝑂𝑃2 to
𝑂𝑃2_𝑛𝑒𝑤, so that the system voltage will maintain at the nominal level.
2.5 Existing control techniques for standalone DC MGs
Among different MG systems (AC or DC, islanded or grid-connected, etc.), this article is aimed to
design a standalone DC microgrid model, and the stable operation of DC voltage of DC bus is the control
target. Thus, a review is done to summarize the existing control techniques for standalone/islanded/off-
grid DC MGs in which DC voltage is the main control objective, as listed in Table 2-3.
Table 2-3. Existing control techniques for standalone DC MGs.
Structure Control objective Proposed method Reference

Centralized DC Voltage LMI Control (Mehdi et al., 2020)
Decentralized DC Voltage and Adaptive Partial Feedback (Mahmud et al.,
Power Dispatch Linearizing Control 2019)
Decentralized DC Voltage Non-fragile 𝐻∞ Control (Hua et al., 2018)
Decentralized DC Voltage Second Order SMC-Based (Cucuzzella et al.,
Control 2018)
Decentralized DC Voltage Nonlinear Integral (Armghan et al.,

Backstepping-Based Control 2020)
Decentralized DC Voltage Fully Decentralized (Amiri et al., 2020)
Backstepping Control
Decentralized DC Voltage LP Based Chebyshev (Lucas-Marcillo et
Theorem al., 2019)
2.5.1 H∞-based control methods
A CC for DC voltage regulation in DC microgrids is proposed in terms of linear matrix
inequalities(LMI). A robust centralized controller is proposed for a DC microgrid impacted by parametric
uncertainty and disturbances(Mehdi et al., 2020). Moreover, the DC voltage control based on the
nonfragile 𝐻∞ method in islanded MGs is investigated under the circumstances of Energy Internet(EI),
and this robust controlled system can be against disturbances and uncertainties(Hua et al., 2018).
Zixuan HE, 936370

15
2.5.2 State-space control methods
A non-linear decentralised control scheme is proposed in order to achieve the ideal voltage on the DC
bus and to maintain the power balance. In this MG model, a PV unit, a fuel cell system, and a BESS are
connected along with DC loads(Mahmud et al., 2019).
2.5.3 Sliding mode control-based control methods
A fully decentralized SMC method is developed to control the DC voltage of converters in DC MGs. In
this model, a decentralised second-order SMC control method is applied to a boost converter, and the
proposed controller could generate a continuous signal as the duty cycle to regulate the DC voltage. The
authors evaluate the feasibility of this method under modelling uncertainties and unknown load
demand(Cucuzzella et al., 2018).
2.5.4 Backstepping-based control methods
An integral backstepping control(IBC) method for DC MGs that includes solar unit, wind unit, and
HESS is introduced with an objective to integrate several RESs on a typical DC bus, and its performance
is compared with other control schemes. The controller is designed to adjust the DC bus voltage for wind,
PV, and HESSs(Armghan et al., 2020). And a fully robust backstepping control scheme is proposed to
control the DC voltage of PV systems in off-grid DC MGs(Amiri et al., 2020). This IBC control method
takes the overall disturbances into using a disturbance observer. In addition, the measurement of local
parameters is integrated for quicker response in tracking the reference signals.
2.5.5 Other methods
As for emerging methods, a control scheme that combines linear programming(LP) and Chebyshev
theorem is proposed to regulate the DC voltage in islanded DC microgrids. This method integrates an
efficient robust controller to solve the problem that happened in LMI by applying the Chebyshev theorem
to the LP approach and in other to reduce the unstable conditions of constant power loads(Lucas-Marcillo
et al., 2019).
Zixuan HE, 936370

16
2.6 MPPT controller for wind-solar system
With the rapid penetration of wind and solar energy in the MGs, it is of great importance to extract the
maximum available power from the wind generators and solar panels to operate the WT and PV systems at
their optimal working conditions. Hence, MPPT algorithms are proposed to make the best use of RESs
and maintain the MGs at maximum output power(Kumar and Chatterjee, 2016).
2.6.1 MPPT algorithm for wind unit
Output power generated from the WTs can be regulated to work efficiently within a specific wind
speed(𝑉𝑤 ) range bounded by cut-in(𝑉𝑤_𝑖𝑛 ) and cut-out(𝑉𝑤_𝑜𝑢𝑡 ) speeds. As shown in Fig. 2-11, the turbine
is stopped and disconnected from the DC bus in the first and fourth region. The second region is between
𝑉𝑤_𝑖𝑛 and 𝑉𝑟𝑎𝑡𝑒𝑑 , where an MPPT controller starts to operate to extract energy. Then in the third region,
between 𝑉𝑟𝑎𝑡𝑒𝑑 and 𝑉𝑤_𝑜𝑢𝑡 , the mechanical power generation is maintained at the rated power(𝑃𝑟𝑎𝑡𝑒𝑑 ) to
avoid damage to WTs(Bianchi, de Battista and Mantz, 2007).
Fig. 2-11. MPPT for WT system based on speed regions.
The MPPT algorithms for wind modules are broadly categorized into indirect power controller(IPC)
and direct power controller(DPC)(Kazmi et al., 2011). The IPC maximizes the captured mechanical wind
power(𝑃𝑤𝑖𝑛𝑑 ), whereas the DPC directly maximizes output electrical power(𝑃𝑜 ). The relationship between
𝑃𝑤𝑖𝑛𝑑 and 𝑃𝑜 is given as
𝑃𝑜 = 𝜂𝑔 𝜂𝑐 𝑃𝑤𝑖𝑛𝑑 (2 − 3)
Zixuan HE, 936370

17
𝜂𝑔 and 𝜂𝑐 refer to generator efficiency and converter efficiency respectively, varying with rotor
𝑜𝑝𝑡
speed(𝜔𝑚 ). Thus, even when the optimal wind speed(𝜔𝑚 ) is obtained, it cannot guarantee the optimal
𝑜𝑝𝑡
electrical power (𝑃𝑜 ).
Table 2-4. Existing MPPT algorithms for wind unit.
MPPT Type Control Logic Reference

High climb DPC Disturb a reference rotational speed (𝜔𝑟𝑒𝑓 ) in and (Wang and
search(HCS) observe the resulting changes in net power curve until Chang, 2004)
the slope becomes zero.
Incremental DPC An optimum DC link voltage (𝑉𝑟𝑒𝑓 ) is defined to (Mosa et al.,

conductance maximize output power of the generator until the slope 2012)
(INC) of P-V characteristic should be zero.
Optimal relation DPC The MPP is tracked by working on the optimal current (Carrillo et al.,
based(ORB) curve at all times based on a pre-obtained curve 2013)
Tip speed IPC Keep λ at an optimum value at which extracted power is (Thongam and
ratio(TSR) maximized by regulating the 𝜔𝑚 Mohand, 2011)
Optimal IPC Control 𝑇𝑚 to obtain 𝑇𝑜𝑝𝑡 reference curve according to (Abdullah et al.,
torque(OT) 𝑃𝑚_𝑀𝐴𝑋 at a given 𝑉𝑤 2012)
Power signal IPC Use a 𝑃 − 𝑉 curve or the WT power expression to get (I. K. Buehring
feedback(PSF) the 𝑃𝑚𝑎𝑥 and Freris,
1981)
2.6.2 MPPT algorithm for PV unit
Since the I–V characteristics of a PV panel strongly depend on solar radiation( 𝐺 ) and cell
temperature(𝑇), the MPP position continuously varies(Nema, Nema and Rangnekar, 2009). To cope with
this problem, specific circuits, called maximum power point trackers(MPPT) are implemented by inserting
a DC/DC converter between the PV modules and the load to ensure that the PV module operates at
MPP(Piegari and Rizzo, 2010). Most MPPT controllers for PV system generate the best duty cycle(𝐷)
based on the voltage&current measurements so that the power remains at MPP (Coelho, Concer and
Martins, 2010). As shown in Fig. 2-12, at the MPP (𝑉𝑚𝑝𝑝 , 𝐼𝑚𝑝𝑝 ), the PV module operates at maximum
Zixuan HE, 936370

18
efficiency and produces its maximum output power (𝑃𝑚𝑎𝑥 ) given by (Bendib, Belmili and Krim, 2015):
𝑃𝑚𝑎𝑥 = 𝑉𝑚𝑝𝑝 × 𝐼𝑚𝑝𝑝 (2 − 4)
Fig. 2-12. I–V and P–V characteristics of a PV module.
MPPT algorithms for PV modules can be broadly categorized into conventional and intelligent methods.
The Intelligent methods are emerging ideas like fuzzy logic(FL) and artificial neutral logic(ANN)(Salas et
al., 2006). Among traditional methods, P&O algorithm and INC algorithm are also most utilized and have
also been long experimented and applied to commercial utilization.
Table 2-5. Existing MPPT algorithms for solar unit.
MPPT Type Control Logic References

P&O Traditional Perturb the operating voltage to observe the (Locment,
power variation so that the direction of voltage Sechilariu and
change can be deduced to give a 𝑉𝑟𝑒𝑓 Houssamo,
2010)
INC Traditional Detect the operating voltage where the (Liu et al.,
instantaneous conductance is equal to the 2007)
incremental conductance based on 𝑃– 𝑉 curve
FL Advanced Three stages: Fuzzification, Inference, (Messai et al.,

Defuzzification 2011)
ANN Advanced Neurons are interconnected by a large number (Sabir, Abd

of weighted links. Each neuron receives Ghani and Abd
voltage and current signals and produces a Elhamid, 2015)
single output power signal.
Zixuan HE, 936370

19
2.7 Summary
In brief, this chapter focuses on a literature review on MG systems. The standalone MGs are first
introduced in terms of system configuration, challenges and proposed solutions. Then EM strategies are
presented to regulate the power system intelligently and orderly, followed by the basic droop control
technique for DC/DC converters that are widely utilized in power systems. Finally, existing control
approaches for DC MG systems operated in standalone mode are summarized, and MPPT algorithms
applied in wind-solar hybrid HRESs are introduced to lay a basis for the model proposed in this article.
Zixuan HE, 936370

20
Chapter 3 Mathematical Modelling of DC MGs Components
3.1 Introduction
In chapter 3, the mathematical modelling of DC microgrid components: DC/DC converters, wind
generators, PV panels, and LIBs are presented respectively based on mathematical derivation. In addition,
the modelling structures of each subsection of the DC microgrid are also presented.
3.2 DC/DC converter modelling
DC/DC converters can be divided into isolated and direct converters depending on whether they have a
transformer or not. In these circuits, transistors and diodes are used to supply switching power, and
inductors(𝐿) and capacitors(𝐶) are used as energy storage elements for energy transfer(Lynser et al., 2018).
This section demonstrates two fundamental direct converter circuits: buck converter and boost converter.
3.2.1 Buck converter
A typical buck converter circuit consists of a switching transistor and a flywheel circuit. The switch
operates at the rate of PWM switching frequency. The voltage at the supply is 𝑉𝑆 and the output voltage is
𝑉0 , the switching period of the switching transistor is 𝑇𝑆 and the duty cycle is 𝐷 (Learn about electronics,
2020). The mathematical explanation is presented as followed.
When 𝑡 = 0, the switching transistor turns on for the duration 𝑇𝑜𝑛 , during which time the diode is
reverse biased and the power supply charges the inductor.

𝑑𝐼𝐿
𝐿 = 𝑉𝑠 − 𝑉0 (3 − 1)
dt
During the time 𝑇𝑂𝑁 , the increase in inductance current(𝐼𝐿 ) is:
𝑉𝑠 − 𝑉0
Δ𝐼𝐿(+) = × 𝑇𝑂𝑁 (3 − 2)
𝐿
When 𝑇 = 𝑇𝑂𝑁 , the switch is turned off for the duration of 𝑇𝑂𝐹𝐹 , during which the diode is switched on
and the inductor is discharged.

𝑑𝐼𝐿
𝐿 = −𝑉0 (3 − 3)
dt
Zixuan HE, 936370
21
During the time 𝑇𝑂𝐹𝐹 , the inductor current is reduced by:

𝑈0
Δ𝐼𝐿(−) = × (𝑇𝑠 − 𝑇𝑂𝑁 ) (3 − 4)
𝐿
According to the energy conservation law (Δ𝑖𝐿(+) = Δ𝑖𝐿(−) ), it can be concluded that:
𝑉𝑠 − 𝑉0 𝑉0
× 𝐷𝑇𝑠 = × (1 − 𝐷)𝑇𝑠 (3 − 5)
𝐿 𝐿
𝑈0 = 𝐷𝑈𝑠 (3 − 6)
Since the duty cycle 𝐷 is a constant of less than 1, the circuit can act as a step-down voltage.
(a) During “ON” Period.
(b) During “OFF” Period.
Fig. 3-1. Working condition of a buck converter.
As shown in Fig. 3-1(a), when the switching transistor is switched on, current flows through the load
via the inductor. Meanwhile, the inductor stores energy from the load with increasing input. During the
‘ON’ period, the charge on the capacitor accumulates gradually, and the increasing current in the circuit
will lead to a large positive voltage on the diode cathode so that it will be reverse biased. Conversely,
when the transistor is switched off, as shown in Fig. 3-1(b), the energy stored in the induction is sent back
into the circuit to keep the current flowing until the transistor switch is open.
Zixuan HE, 936370

22
3.2.2 Boost converter
The boost converter converts a low input voltage to a high output voltage, functioning like a reversed
buck converter(Ustun and Mekhilef, 2010). A basic boost converter circuit consists of a switching
transistor, the rest of the components are the same as those used in the buck converter mentioned
above(Learn about electronics, 2020).
When 𝑡 = 0, the switching transistor is switched on for the duration 𝑇𝑂𝑁 , during which the diode is
switched off and all the power from the supply is used to charge the inductor.
𝑑𝐼𝐿
𝐿 = 𝑉𝑠 (3 − 7)
dt
During the time 𝑇𝑂𝑁 , the increase in inductance current is:
𝑉𝑠
Δ𝐼𝐿(+) = × 𝑇𝑜𝑛 (3 − 8)
𝐿
When 𝑇 = 𝑇𝑂𝑁 , the transistor is switched off for the duration of 𝑇𝑂𝐹𝐹 , during which the diode is
switched on and both the power supply and inductor supply the load and charge the capacitor as well:
𝑑𝑖𝐿
𝐿 = 𝑉𝑠 − 𝑉0 (3 − 9)
dt
During the time 𝑇𝑂𝐹𝐹 , the inductor current is reduced by:
𝑉0 − 𝑉𝑠
Δ𝑖𝐿(−) = × (𝑇𝑠 − 𝑇 𝑂𝑁 ) (3 − 10)
𝐿
According to the energy conservation law (Δ𝑖𝐿(+) = Δ𝑖𝐿(−) ), it can be concluded that:
𝑉𝑠 𝑉0 − 𝑉𝑠
× 𝐷𝑇𝑠 = × (1 − 𝐷)𝑇𝑠 (3 − 11)
𝐿 𝐿
1
𝑉0 = (3 − 12)
(1 − 𝐷)𝑉𝑠
Since the duty cycle 1 − 𝐷 is also a constant less than 1, the circuit can act as a voltage booster.
(a) Step 1. (b) Step 2.
Zixuan HE, 936370

23
(c) Step 3.
Fig. 3-2. Working condition of a boost converter.
Fig. 3-2 shows the operating logic of a basic boost converter. In step 1, as shown in Fig. 3-2(a), the
current flows through the inductor when the switch is closed, and the inductor stores energy in the
magnetic field. In step 2, as shown in Fig. 3-2(b), after the transistor is switched off, the inductor generates
a large voltage opposite to the original voltage supplied to it using the energy stored previously. In order to
maintain a stable current flow, the capacitor stays charged since it cannot discharge through the reverse-
biased diode. Then in step 3, as shown in Fig. 3-2(c), the diode is switched off because the cathode of the
diode is more positive than its anode, thus the load is disconnected with the DC input. Nonetheless, the
capacitor continues to supply the load with the charge stored when the transistor switches off, thereby
maintaining a steady output voltage for the load.
3.3 WT system modelling
Fig. 3-3. Model of WT system.
As illustrated in Fig. 3-3, the WT system simulated in this article consists of a WT module together with
Zixuan HE, 936370

24
the permanent magnet synchronous generator(PMSG), a diode rectifier, and a boost converter. Instructed
by an MPPT controller, the wind generator can be switched between MPPT or off-MPPT mode.
The input of a WT system is wind speed and the output is the mechanical power generated by the
generator rotor, which is connected to wind turbines by a gearbox setup. The WT system can convert the
wind energy into unstable alternating current, then convert the alternating current into direct current
through the rectifier(Baroudi, Dinavahi and Knight, 2007).
According to Betz’s Law, the output mechanical power and torque developed by a WT are given by:
1
𝑃𝑚 = ∙ 𝐶 ∙ 𝜌 ∙ 𝐴 ∙ 𝑉𝑤 3 (3 − 13)
2 𝑝
1 3
𝑃𝑚 2 ∙ 𝐶𝑝 ∙ 𝜌 ∙ 𝐴 ∙ 𝑉𝑤
𝑇𝑚 = = (3 − 14)
𝜔𝑚 𝜔𝑚
𝑃𝑚 is the turbine output power( 𝑊 ); 𝑇𝑚 is the mechanical torque of turbine( 𝑁 ∙ 𝑚 ); 𝜌 is the air
density(𝑘𝑔/𝑚3); 𝐴 is the area of turbine blades(𝑚2 ); 𝑅 is the radius of the turbine blade(𝑚); 𝑉𝑤 is the
wind speed(𝑚/𝑠); 𝜔𝑚 is the mechanical angular speed of the turbine(𝑟𝑎𝑑/𝑠).
𝐶𝑝 is the turbine power coefficient(𝑑𝑖𝑚𝑒𝑛𝑠𝑖𝑜𝑛𝑙𝑒𝑠𝑠), it is defined by a function of blade tip speed ratio
(𝜆) and blade pitch angle (𝛽) according to specific WT parameters(Ragheb, 2014):
1 1
(−𝑘 )
𝐶𝑃 (𝜆, 𝛽) = 𝑘1 (𝑘2 − 𝑘3 𝛽 − 𝑘4 𝛽𝑘5 − 𝑘6 ) exp 7 𝜆𝑖 (3 − 15)
𝜆𝑖
1 1 0.035
= − (3 − 16)
𝜆𝑖 𝜆 + 0.08𝛽 1 + 𝛽 3
𝑅𝜔𝑚
𝜆= (3 − 17)
𝑉𝑤
The mechanical angular speed(𝜔𝑚 ) of WTs can be derived by the relation between electrical and
mechanical torque:
𝑑𝜔𝑚
𝐽 = 𝑇𝑒 − 𝑇𝑚 − 𝐵 ⋅ 𝜔𝑚 (3 − 18)
𝑑𝑡
𝑇𝑒 is the electromagnetic torque(𝑁 ∙ 𝑚), 𝐽 is the combined inertia of the WTs and rotor(𝑁 ∙ 𝑚2 ), and 𝐵
refers to the viscous friction of the rotor (𝑁 ∙ 𝑚 ∙ 𝑠/𝑟𝑎𝑑). Using the equations obtained from (3 − 13) to
(3 − 18).
The dynamic model of the wind generator unit can be built as shown in Fig. 3-4.
Zixuan HE, 936370

25
Fig. 3-4. Dynamic model of the wind generator.
3.4 PV system modelling
As illustrated in Fig. 3-5, the PV system simulated in this article consists of a PV module, a controller
current source(CCM) that converts input signals(temperature and irradiance) into an equivalent current
source, a load, and a boost converter. Like the WT system, the solar generator can be switched between
MPPT or off-MPPT mode by an MPPT controller.
Fig. 3-5. Model of PV system.
In a photovoltaic cell, the input is solar radiation and temperature, and the output refers to generated
power which is the product of voltage and current(𝑃 = 𝑉 × 𝐼). A PV cell consists of a p–n junction
integrated in a thin layer of semiconductor, and their characteristics are similar to p–n diodes(Asim et al.,
2012). As shown in Fig. 3-6, a solar cell can be represented by a single-diode electrical circuit(Pradeep et
al., 2012).
Zixuan HE, 936370

26
Fig. 3-6. Equivalent electrical circuit of a solar cell.
According to KCL and KVL laws,
𝐼𝑝𝑣 + 𝐼𝑑 + 𝐼𝑠ℎ = 𝐼𝑝ℎ (3 − 19)

𝑉𝑝𝑣 + 𝐼𝑝𝑣 × 𝑅𝑠
𝐼𝑠ℎ = (3 − 20)
𝑅𝑠ℎ
According to the characteristics of diode,
𝑞(𝑉𝑝𝑣 + 𝐼𝑝𝑣 𝑅𝑠 )
𝐼𝑑 = I0 [exp ( ) − 1] (3 − 21)
𝐴𝑘𝑇
Since the output power from a single PV cell is small, PV cells are connected in series and parallel to
generate the required voltage and power(Ishaque, Salam and Syafaruddin, 2011). Thus, the I–V
characteristic equation of a PV module arranged in 𝑁𝑃 parallel and 𝑁𝑆 series solar cell can be expressed
by substituting formula (2-20) and (2-21) into (2-19), yields:

𝑞(𝑉𝑝𝑣 + 𝐼𝑝𝑣 × 𝑅𝑠 ) 𝑉𝑝𝑣 + 𝐼𝑝𝑣 × 𝑅𝑠
𝐼𝑝𝑣 = 𝐼𝑝ℎ − 𝐼𝑜 [exp ( ) − 1] − (3 − 22)
𝐴𝑘𝑇 𝑅𝑠ℎ
𝐼𝑝𝑣 and 𝑉𝑝𝑣 are the output current and output voltage of the photovoltaic cell; 𝑅𝑆 and 𝑅𝑃 is the series
and parallel resistance; 𝐴 is the diode ideality factor; 𝑘 is the Boltzmann constant (1.38 × 10−23 J/K), q is
the electron charge (1.67 × 10−19 𝐶); T is the absolute temperature of solar panels; 𝐼𝑜 is the diode’s
reverse saturation current; 𝐼𝑝ℎ is the generated photocurrent, which is influenced by the irradiation and
temperature.
In practical applications, manufacturers generally only provide the following parameters: 𝐼𝑠𝑐 (short-
circuit current), 𝑉𝑜𝑐 (open-circuit voltage), 𝐼𝑚 (current for MPP), 𝑉𝑚 (voltage for MPP), so as the PV
module in the MATLAB library. Moreover, 𝑅𝑠ℎ is far larger than 𝑅𝑠 , and the resistance of the forward-
biased diode is also larger than 𝑅𝑠ℎ . Hence, formula (3-22) can be simplified as
Zixuan HE, 936370

27
𝑉𝑝𝑣
𝐼𝑝𝑣 = 𝐼sc − 𝐼sc 𝐶1 [exp ( ) −1] (3 − 23)
𝐶2 𝑉𝑜𝑐
Solve the equation (11) when operating in open circuit: 𝐼𝑝𝑣 = 0, 𝑉𝑝𝑣 = 𝑉𝑜𝑐 ; when operating at MPP:
𝑉𝑝𝑣
( )
𝐼𝑝𝑣 = 𝐼𝑚 , 𝑉𝑝𝑣 = 𝑉𝑚 . At room temperature, 𝑒 𝐶2 𝑉𝑜𝑐 is much larger than 1, which can be neglected. Hence,
𝐶1 and 𝐶2 can be deduced as:

𝐼𝑚 𝑉𝑚
𝐶1 = (1 − ) 𝑒𝑥𝑝 (− ) (3 − 24)
𝐼𝑠𝑐 𝐶2 𝑉𝑜𝑐
𝑉
(𝑉𝑚 − 1)
𝑜𝑐
𝐶2 = (3 − 25)
𝐼
𝐼𝑛 (1 − 𝐼𝑚 )
𝑠𝑐
Equations (2-23) to (2-25) are derived based on a mathematical model under standard environmental
conditions: (𝑇𝑟𝑒𝑓 = 25∘ C, 𝐺𝑟𝑒𝑓 = 1000 W/m2 ). However, since temperature(𝑇) and irradiance(𝐺) have
a huge impact on the solar cell, it is necessary to make corrections when the PV panel is not operating
under standard conditions. Hence, with correction parameters 𝐷1 and 𝐷2 , the actual PV current can be
expressed as:
′ ′
𝑉𝑝𝑣
𝐼𝑝𝑣 = 𝐼𝑠𝑐 − 𝐼𝑠𝑐 𝐶1 [exp ( ) − 1] (3 − 26)
𝐶2 𝑉𝑜𝑐′
′
𝐼𝑠𝑐 = 𝐼𝑠𝑐 𝐷1
′
𝐼𝑚 = 𝐼𝑚 𝐷1
(3 − 27)
𝑉𝑠𝑐′ = 𝑉𝑠𝑐 𝐷2
{ 𝑉𝑚′ = 𝑉𝑚 𝐷2
3.5 ESS module modelling
Fig. 3-7. Model of ESS module.
Zixuan HE, 936370

28
Fig. 3-7 shows the ESS connected to the DC bus. The LIB model is directly called from the MATLAB
Library. It is connected to the DC bus via a bidirectional DC-DC converter, which consists of an inductor
𝐿𝑏𝑎𝑡 , resistance 𝑅𝑏𝑎𝑡 , filter capacitor 𝐶𝑏𝑎𝑡 , and two switches 𝑆3 and 𝑆4 . The converter can be operated in
charging or discharging mode to regulate the DC bus voltage.
LIB technology is based on lithium-intercalation compounds. As shown in Fig. 3-8, the lithium ions
travel between the anode and cathode via the electrolyte(Thackeray, Wolverton and Isaacs, 2012). During
the discharge process, cathode and anode undergo oxidation and reduction reactions, lithium ions(𝐿𝑖 + ) are
released from the anode and diffused into the de-lithiated cathode(Levi, Gofer and Aurbach, 2010).
Fig. 3-8. Working principle of LIBs in (a) Charging condition; (b) Discharging condition.
During the charging process, electrons on the cathode flow to the anode electrode via the external
circuit. Through the curved hole on the diaphragm in the electrolyte, 𝐿𝑖 + diffuses from the positive
electrode(cathode) to the positive electrode(anode). where they combine with the electrons that have been
arrived before. Chemical equations for reaction on anode and cathode are shown as followed:
𝐂𝐚𝐭𝐡𝐨𝐝𝐞: LiCoO2 ⟶ Li1−𝑥 CoO2 + 𝑥Li+ + 𝑥𝑒 − (3 − 28)
𝐀𝐧𝐨𝐝𝐞: 6C + xLi+ + xe ⟶ Lix C6 (3 − 29)
During the discharging process, both electrons and 𝐿𝑖 + flow simultaneously with the same direction but
different paths. With a resistor that allows electrons to pass, electrons travel from the anode electrode
through the electron conductor to the cathode electrode, and 𝐿𝑖 + diffuses into the electrolyte from the
anode electrode, flows through the diaphragm, and combines with the electrons that have already been
reached.
Here are the characteristics parameters of LIB sector:
(1) Charging capacity (𝑄𝑐 ): the amount of power received by the battery, as shown in equation (3-30):
Zixuan HE, 936370

29
𝑡𝑐
𝑄𝑐 = ∫ 𝐼𝑐 𝑑𝑡 (3 − 30)
0
(2) Discharge capacity (𝑄𝑑 ): the amount of electricity discharged from the battery when fully charged,
as shown in equation (3-31):

𝑡𝑑
𝑄𝑑 = ∫ 𝐼𝑑 𝑑𝑡 (3 − 31)
0
Where 𝐼𝑐 and 𝐼𝑑 refer to the charging and discharging currents respectively; 𝑡𝑐 and 𝑡𝑑 represent the
charging and discharging times respectively.
(3) State of charge(𝑆𝑜𝐶): the residual charge of the battery, is expressed as a ratio of the current
residual capacity to the fully charged state, as shown in equation (3-21):

𝑄(𝑡)
𝑆𝑂𝐶 = (3 − 32)
𝑄𝑏
(4) Depth of discharge(𝐷𝑜𝐷):
𝐷𝑂𝐷 = 1 − 𝑆𝑂𝐶 (3 − 33)
3.6 Summary
In short, chapter 3 first introduces the two basic switching mode DC/DC converters: buck converter and
boost converter. Based on the operating logic of power converters, the modelling structures of WT system
and PV system, and ESS module are demonstrated subsequently with an integration of a boost type
DC/DC converter and other electronics components; and the converter circuit of LIB sector is designed
via a bidirectional DC/DC converter. And the working principles of wind generators, solar panels and LIB
modules are explained respectively according to the proposed theory and mathematical derivation.
Zixuan HE, 936370

30
Chapter 4 MPPT algorithm
4.1 Introduction
In chapter 3, the available MPPT algorithms utilized in the model or as control groups are described for
extracting maximum power from wind generators and solar panels. Among the most mainstream MPPT
controllers, Perturb and Observation(P&O) algorithm (so-called HCS algorithm in WT systems) and
Incremental conductance(INC) MPPT algorithm are widely used both in WT and PV systems to track the
MPP. After a brief review of MPPT algorithms in chapter 2, this chapter explains how P&O and INC
method controls the power converter in WT and PV systems.
4.2 MPPT algorithm for WT system
Fig. 4-1. MPPT principles based on variable wind speeds.

Since the air density(𝜌) and turbine parameters are constant values, the turbine output power(𝑃𝑚 )
depends on the rotor speed of WT(ω𝑚 ), namely the value of 𝐶𝑃 , Therefore, 𝐶𝑃 would be maximized when
𝜔𝑚 is maximal. As shown in Fig. 4-1, the output power (𝑃𝑚 ) can be controlled by adjusting rotor
speed(𝜔𝑚 ) so that the 𝐶𝑃_𝑀𝐴𝑋 can be instantaneously captured and maintained at optimal power point.
4.2.1 P&O Algorithm for WT system
The Perturb and Observation(P&O) algorithm, also named HCS algorithm in WT systems, is a DPC
MPPT algorithm that tracks the operating power point based on a pre-obtained system curve(I. K.
Zixuan HE, 936370

31
Buehring and Freris, 1981). The tracking process is realized by an increment or decrement of reference
𝑟𝑒𝑓
rotational speed(𝜔𝑚 ) on the net power curve, then the operating power is adjusted to approach the MPP
until the slope becomes zero(Koutroulis and Kalaitzakis, 2006).
As shown in Fig. 4-2(a), when the operating point is on the left side of the MPP, namely the net
𝑜𝑝𝑡
power(𝑃𝑚 ) is deviated from 𝑃𝑚 on the left side, the MPPT controller is activated to adjust 𝑃𝑚 to the
𝑜𝑝𝑡
right with an increment of rotor speed(∆𝜔𝑓𝑖𝑥𝑒𝑑 ) to approach 𝑃𝑚 , and vice versa on the right side.
Fig. 4-2. P&O method with (a) fixed step; (b) adaptive step.
However, fixed step change could not deal with rapid wind variation, which could result in a wrong
direction detection to reach MPP(Wang and Chang, 2004). To overcome this problem, an adaptive step
𝑜𝑝𝑡
size(𝜔𝑎𝑑𝑝 ) is introduced so that the distance between the operating point and 𝑃𝑚 can be intelligently
observed, as shown in Fig. 4-2(b). Compared with the fixed step, the adaptive step is relatively smaller,
leading to a slower initial starting time, whereas the value of 𝛥𝜔𝑎𝑑𝑝 will decrease if the output power is
closer to the MPP so that the oscillation around MPP is reduced(Jinbo et al., 2014).
Fig. 4-3. Flowchart of improved P&O MPPT algorithm.
Zixuan HE, 936370

32
Furthermore, since the air density also varies depending on weather conditions, the MPP curve is
characterized by 𝑃𝑜𝑝𝑡 = 𝑘𝑜𝑝𝑡 𝜔3 also changes(Peng, Feng and Song, 2013). Hence, the improved
algorithm is proposed to combine two parts. In the first part, P&O algorithm is utilized to track the MPP
based on the current 𝑘𝑜𝑝𝑡 value. After reaching the MPP, the second part uses the updated 𝑘𝑜𝑝𝑡 to generate
a precise net power curve. As depicted in Fig. 4-3, there are three steps included in the improved P&O
MPPT algorithm. Step 1 searches for a 𝑘𝑜𝑝𝑡 to track MPP using the P&O algorithm. Step 2 maintains the
WT system at the MPP until air density(𝑉𝑤 ) varies. Then, the adjusted 𝑘𝑜𝑝𝑡 is updated in step 3 for the
following tracking process(Wang and Chang, 2004).
4.2.2 INC algorithm for WT system
Incremental conductance(INC) algorithm is also a DPC-based MPPT algorithm in which the MPP is
tracked by detecting the output power(𝑃𝑑𝑐 ) of the rectifier(Kumar and Chatterjee, 2016). The output
power(𝑃𝑑𝑐 = 𝑉𝑑𝑐 × 𝐼𝑑𝑐 ) of WTs can be represented as:

𝑑𝑃𝑑𝑐 𝑑(𝑉𝐼)𝑑𝑐 𝑑𝐼𝑑𝑐
= = 𝐼𝑑𝑐 + 𝑉𝑑𝑐 (4 − 1)
𝑑𝑉𝑑𝑐 𝑑𝑉𝑑𝑐 𝑑𝑉𝑑𝑐
𝐼𝑑𝑐 𝑑𝐼𝑑𝑐
+ =0 (4 − 2)
𝑉𝑑𝑐 𝑑𝑉𝑑𝑐
As shown in Fig. 4-4, an optimum DC voltage (𝑉𝑟𝑒𝑓 ) is defined to maximize output power of the wind
generator using the INC MPPT algorithm. The tangent slope of the P-V characteristic is equal to zero at
MPP, positive on the left side, and negative on the right side of MPP.
Fig. 4-4. Operating principle of INC method in WT system.
Similar to P&O algorithm, the modified INC algorithm improves the tracking performance by
Zixuan HE, 936370

33
integrating an adaptive step for 𝑉𝑑𝑐 variations, the integration of predictive control module in WT MPPT
controller improves both tracking speed and power reliability. As shown in Fig. 4-5, the step size is
intelligently adjusted to track the MPP. Compared with fixed step variations (𝑉𝑑𝑐_𝑓𝑖𝑥𝑒𝑑 ), a modified INC
INC algorithm causes the operating point to reach the MPP more rapidly and reduces the oscillations
around MPP, which leads to higher accuracy and quicker convergence speed(Mosa et al., 2012).
Fig. 4-5. Flowchart of INC algorithm with adaptive step in WT system.
4.3 MPPT controllers for PV system
Fig. 4-6. I–V characteristics of a PV module under (a) Changing irradiance at 𝟐𝟓℃; (b) Changing
temperature at 𝟏𝟎𝟎𝟎𝑾/𝒎𝟐 .
As shown in Fig. 4-6, the variation in solar irradiance has an obvious impact on the PV current(𝐼𝑝𝑣 ), but
the PV voltage(𝑉𝑝𝑣 ) stays almost constant. Alternatively, the voltage varies under a changing temperature
Zixuan HE, 936370

34
while the current remains almost unchanged.
Fig. 4-7 confirms how the PV current(𝐼𝑝𝑣 ) and PV voltage(𝑉𝑝𝑣 ) changes under the variation of solar
irradiance(𝐺) and cell temperature(𝑇) by changing the Y-axis to output power(𝑃𝑝𝑣 ). Fig. 4-7(a) shows that
the output power(𝑃𝑝𝑣 ) of the PV module decreases by an irradiance decrement. On the other hand, Fig. 4-7
(b) shows that the output power(𝑃𝑝𝑣 ) reduces by an increment of cell temperature.
Fig. 4-7. P–V characteristics of a PV module under (a) Changing irradiance at 𝟐𝟓℃; (b)
Changing temperature at 𝟏𝟎𝟎𝟎𝑾/𝒎𝟐 .
4.3.1 P&O algorithm for PV system
Fig. 4-8. (a) Flowchart of P&O MPPT algorithm; (b) Path of MPP under changing irradiance.
The P&O method is widely employed in PV based HRESs due to its low-cost, simplicity and easy
implementation(Femia et al., 2005). As shown in Fig. 4-8(a), if the power of the PV module increases for
a given direction of operating voltage perturbation, it can be concluded that the operating power is moving
Zixuan HE, 936370

35
towards the MPP, thus a further perturbation must be done to the operating voltage in the same direction as
the last time. Conversely, if the power of the PV module decreases, it indicates the operating point is
moving away from the MPP, so the perturbation direction of operating voltage should be reversed.
However, as shown in Fig. 3.10, the power of PV module also changes because of the changing weather
conditions rather than voltage perturbation(Femia et al., 2013). To overcome this issue, an improved P&O
algorithm takes the variations of power (𝛥𝑃) and voltage (𝛥𝑉) at consecutive time intervals (𝑘 − 1) and (𝑘)
into consideration, and then uses four variables (𝛥𝑉(𝑘 − 1), 𝛥𝑃(𝑘 − 1), 𝛥𝑉 (𝑘) 𝑎𝑛𝑑 𝛥𝑃(𝑘)), generating
16 potential circumstances, as shown in Table 4-1. If the increase in power is caused by the voltage
perturbation rather than the increment of irradiance, the MPPT controller causes a decrease in power for
one scanning cycle (wrong direction of perturbation) but takes the correct direction again in the next
scanning cycle(Kakosimos and Kladas, 2011).
Table 4-1. Truth table for improved P&O algorithm.
Case 𝛥𝑉(𝑘 − 1) 𝛥𝑃(𝑘 − 1) 𝛥𝑉 (𝑘) 𝛥𝑃(𝑘) 𝛥𝑉(𝑘 + 1)

1 − − − − +
2 − − − + +
3 − − + − −
4 − − + + +
5 − + − − +
6 − + − + +
7 − + + − −
8 − + + + −
9 + − − − +
10 + − − + −
11 + − + − −
12 + − + + −
13 + + − − +
14 + + − + +
15 + + + − −
16 + + + + −
4.3.2 INC algorithm for PV system
The incremental conductance (INC) algorithm is similar to the INC algorithm mentioned in WT system.
It is based on the slope of the solar power versus voltage (𝑃– 𝑉) curve(Liu et al., 2007). The aim of INC
algorithm is to find the voltage operating point where the instantaneous conductance (𝐼/𝑉) is equal to the
Zixuan HE, 936370

36
incremental conductance (𝑑𝐼/𝑑𝑉).
Fig. 4-9. Operating principle of INC algorithm in PV system.
In PV systems, the INC controllers monitor the output voltage and current, which are used to calculate
the simultaneous conductance(Liu et al., 2008). And the results are served as criteria to decide whether an
increase or decrease in the output duty cycle. The output power can be expressed as:
𝑑𝑃 𝑑(𝐼𝑉) 𝑑𝐼 𝑑𝑃 1 𝑑𝐼 𝐼
= =𝑉 + 𝐼 ====⇒ × = + (4 − 3)
𝑑𝑉 𝑑𝑉 𝑑𝑉 𝑑𝑉 𝑉 𝑑𝑉 𝑉
𝑃, 𝑉, 𝑎𝑛𝑑 𝐼 refer to output power, voltage, and current, respectively. As shown in Fig. 4-9, the slope of
the power curve is equal to zero at the MPP, increases to the left of the MPP and decreases to the right of
the MPP. And the three situations can be expressed as:

𝑑𝑃 𝑑𝐼 𝐼
= 0, = − , (𝑎𝑡 𝑀𝑃𝑃) (4 − 4)
𝑑𝑉 𝑑𝑉 𝑉
> 0, > − , (𝑙𝑒𝑓𝑡 𝑜𝑓 𝑀𝑃𝑃) (4 − 5)
< 0, > − , (𝑟𝑖𝑔ℎ𝑡 𝑜𝑓 𝑀𝑃𝑃) (4 − 6)
As shown in Fig. 4-10, operating voltage and current are measured to calculate the incremental changes
of 𝑑𝑉 and 𝑑𝐼 at consecutive time intervals (𝑘 − 1) and (𝑘), which are expressed as
𝑑𝑉(𝑘) = 𝑉(𝑘) − 𝑉(𝑘 − 1) (4 − 7)
𝑑𝐼(𝑘) = 𝐼(𝑘) − 𝐼(𝑘 − 1) (4 − 8)
At the MPP, 𝑉𝑟𝑒𝑓 is equal to 𝑉𝑚𝑝𝑝 . Once the MPP is reached, the PV module is maintained at MPP
unless a change in 𝑑𝐼 is detected. This algorithm changes 𝑉𝑟𝑒𝑓 with an adjustment step width constant (𝐶𝑎 )
Zixuan HE, 936370

37
to track the next MPP(Esram and Chapman, 2007).
Fig. 4-10. Flowchart of INC MPPT algorithm in PV system.
4.4 Summary
In short, this chapter introduces P&O and INC algorithms utilized to control DC/DC converters in WT
and PV systems. Both of the two algorithms are DPC-based methods that directly control the electrical
power regardless of generator efficiency and converter efficiency. And both of them control the output
based on an output power versus voltage curve. Moreover, the idea of adaptive step size is integrated both
in P&O and INC controllers to intelligently detect the perturbed direction or adjust the step size in order to
reduce the tracking error between the operating point and MPP.
As for the improved P&O method, according to formulas (1) and (2), the output of wind generator is
restricted by the air density, hence the WT system can be optimized by introducing a parameter 𝑘𝑜𝑝𝑡 to
dynamically adjust the system curve(𝑃 − 𝑉). Alternatively, the P&O MPPT controller for PV system can
also be improved by supervising the output change because the power of PV module also changes because
of the changing weather conditions rather than voltage perturbation.
Zixuan HE, 936370

38
Chapter 5 Control Techniques for ESS Module
5.1 Introduction
After the review of the MPPT algorithm for WT and PV systems, chapter 5 subsequently demonstrates
the ADRC scheme utilized to control the LIB sector. And a comparison is made between ADRC scheme
and traditional PID based method to regulate DC voltage. Since the operation of DC/DC converter in ESS
directly leads to different working modes of the microgrid, the energy management(EM) approach is also
presented in this chapter.
5.2 ADRC scheme for ESS module
In the ESS module, the ADRC(Active Disturbance Rejection Control) scheme is utilized in the DC/DC
converter to control the DC voltage. The ADRC scheme is a recently proposed control technique that has
the simplicity of traditional PID controllers but uses errors to eliminate errors so that the parameters can
be controlled without relying on an accurate system model(Zhou et al., 2018).
Fig. 5-1. ADRC control structure block diagram.
As shown in Fig. 5-1, a typical ADRC controller is composed of three components: a tracking
differentiator (TD), an extended state observer (ESO), and non-linear state error feedback (NLSEF)
control law(LI, ZHU and LIU, 2012). The core of this method is to consider the internal and external
uncertain disturbances as “generalized disturbances” by TD, try to estimate them in real-time by using the
ESO, and then use it in the NLSEF control law so that the disturbances can be compensated rapidly.
Zixuan HE, 936370

39
Fig. 5-2. ADRC based DC-voltage control structure block diagram.
In ADRC, is used to regulate the DC voltage of the ESS module, as shown in Fig. 5-2. To introduce the
ESO, an augmented state-space formulation is derived by:

𝑥̇ 0 1 𝑥1 𝑏 0
[ 1] = [ ] [ ] + [ 0] 𝑢 + [ ] ℎ (5 − 1)
𝑥̇ 2 0 0 2 𝑥 0 1
where 𝑥1 equals to 𝑉𝑑𝑐 , 𝑥2 equals to the sum of lumped disturbances and uncertainties; 𝑏0 is the of the
input coefficient. ℎ is the derivative signal of the augmented state 𝑥2 .
According to (5-1), the ESO is established as:

𝑥̂̇ 0 1 𝑥̂1 𝑏 𝑙
[ 1] = [ ] [ ] + [ 0 ] 𝑢 + [ 1 ] (𝑦 − 𝑥̂1 ) (5 − 2)
𝑥̂̇2 0 0 𝑥̂2 0 𝑙2
𝑥1
𝑦 = [1 0] [𝑥 ] + 𝛿(𝑡) (5 − 3)
2
where 𝑥̂1 , 𝑥̂2 refer to the estimated state of 𝑥1 and 𝑥2, 𝑦 is the generated output signal which consists of
true signal 𝑥1 and noise signal 𝛿(𝑡), 𝑙1 and 𝑙2 are the ESO gains that can be solved using the bandwidth
tuning method.
𝜆(𝑠) = 𝑠 2 + 𝑙1 𝑠 + 𝑙2 (5 − 4)
By subtracting (5-3) from (5-2) and combining (5-4), the estimated error dynamic can be derived as
follows:
𝑒̇ −𝑙 1 𝑒1 −𝑙 0
[ 1] = [ 1 ] [ ] + [ 1 ] 𝛿(𝑡) + [ ] ℎ (5 − 5)
𝑒̇2 −𝑙2 0 𝑒2 ⏟−𝑙2 1
disturbance term
Hence, the greater the ESO gain is, the better the tracking performance achieved by the ESO.
Zixuan HE, 936370

40
Compared with a traditional PI-based DC-voltage controller, as shown in Fig. 5-3, the output DC
2 ∗ 2
voltage(𝑉𝐷𝐶 ) is compared with the reference signal(𝑉𝐷𝐶 ) to generate the error signal, which is processed
∗ 2
by a voltage compensator(𝐾𝑉𝐶 (𝑠)) and a limiter to give the power reference signal(𝑃𝐷𝑀 ). Finally, the
2
adjusted value of 𝑉𝐷𝐶 generated through the power control loop and DC-link plant(Zhang and Zhou, 2020).
Since the PI-based solutions require detailed characteristics of the DC MG model, the ADRC introduces
an ESO to estimate the interior and exogenous disturbances instead of directly measuring them.
Fig. 5-3. PI-based DC-voltage control structure block diagram.

∗
To conclude, the DC voltage reference(𝑉𝑑𝑐 ) first serves as an input, which is processed by the TD to
give out a generalized disturbance. Secondly, this disturbance signal is sent into the NLSEF module
together with the reference battery current generated in the last scanning cycle, in order to generate the
compensation value for the operating DC voltage. And the ESO module is utilized to observe and estimate
the DC voltage and compensation value in real-time. This module engages in both the inner and outer loop
so that the outcome can be controlled more precisely and instantly in the ways that use errors to eliminate
errors.
5.3 EM Strategy Formulation
According to the power conservation principles, the total power(𝑃𝑛𝑒𝑡 ) balance equation of WT system,
PV system, and ESS module DC microgrid can be expressed as
𝑃𝑛𝑒𝑡 = 𝑃𝑃𝑉 + 𝑃𝑊𝑇 (±𝑃𝐸𝑆𝑆 ) − 𝑃𝑙𝑜𝑎𝑑 (5 − 6)
In this formula, 𝑃𝑛𝑒𝑡 is the sum of each distributed power source and load power in the MG system.
When the sum of WT and PV power is greater than the load power, the total power is greater than zero
(𝑃𝑛𝑒𝑡 > 0), and the ESS module stores excessive energy; when the total power is less than 0 (𝑃𝑛𝑒𝑡 < 0),
Zixuan HE, 936370

41
the ESS module releases the stored energy. With a DC/DC converter, each distributed power source is
connected to the DC bus and is connected in parallel to the DC load. The relation between net power and
DC voltage can be expressed as:

𝑑𝑉𝑑𝑐 1
𝑉𝑑𝑐 = 𝑃 (5 − 7)
𝑑𝑡 𝐶𝑑𝑐 𝑛𝑒𝑡
It can be noticed that maintaining the DC bus voltage at a constant value is a key factor to ensure the
stable power balance of the DC microgrid. Thus, the net power of the DC bus should be maintained at
zero( 𝑃𝑛𝑒𝑡 = 0).
In LIB sector, the bidirectional DC/DC converter is utilized to regulate the DC bus voltage. It can be
operated in either discharging or charging mode depending on the climate condition and load demand. The
working condition of the converter is different when operating at two different modes. These two modes
can be mathematically expressed as:
1, if (𝑖bat_ref > 0)
𝐾={ (5 − 8)
0, if (𝑖bat_ref < 0)
where 𝑖 𝑏𝑎𝑡_𝑟𝑒𝑓 is the battery current reference generated based on the SOC and power demand.
Table. 5-1. Working condition of each DC MG components.

Time(s)
0 1.5 2 2.5 3 4 5
Wind speed(m/s) 8 8 8 6 6 8 8
Irradiance(W/m2) 800 800 1000 1000 600 600 600
𝑖 𝑏𝑎𝑡_𝑟𝑒𝑓 <0 >0 >0 >0 >0 >0 >0
ESS Mode(K) 0 1 1 1 1 1 1
During discharging mode, as shown in Fig. 5-4, 𝑆3 is ON and 𝑆4 is OFF so that the battery provides the
energy to the load via the DC bus (𝑖𝑏𝑎𝑡_𝑟𝑒𝑓 > 0). This mode reflects the circumstances under which solar
panels and wind turbines cannot provide enough for the load, so the battery sector is instructed by the
controller to connect with the DC bus so that the power shortage of the load can be met by ESS system.
Conversely, as shown in Fig. 5-5, if 𝑆3 is OFF and 𝑆4 is ON, the converter is operated at charging mode
(𝑖𝑏𝑎𝑡_𝑟𝑒𝑓 < 0). At this mode, the LIB sector is disconnected from the load to recharge the battery. The
Zixuan HE, 936370

42
current direction is reversed so that the excessive power from WT and PV system is utilized to charge the
battery sector. In other words, this mode reflects the circumstances under which the load power is all
supplied by the WT and PV system, and the ESS is required to store energy for the intermittent period.
Fig. 5-4. Working condition of ESS module at discharging mode.
Fig. 5-5. Working condition of ESS module at charging mode.
5.4 Summary
In this chapter, the ADRC scheme is introduced and programmed to optimize the traditional PI-based
DC voltage scheme using an ESO module that uses errors to estimate and eliminate errors. Furthermore,
based on the converter circuits characteristics that have been mentioned in chapter 2 and 3, the EM
Zixuan HE, 936370

43
strategy is demonstrated by the different modes switched by the bidirectional DC/DC converter in ESS
module so that the DC voltage can be dynamically maintained. Fig. 5-6 presents the whole EM structure
of the DC microgrid: the electricity generated by WT and PV systems is transformed into the required DC
current via DC/DC converters, which are connected in parallel to the DC bus and then recharged or
supplied to the load.
Fig. 5-6. the EM structure of the DC microgrid.
Zixuan HE, 936370

44
Chapter 6 System Simulation and Analysis
6.1 Introduction
The electricity generated by WT and PV systems is transformed into the required DC current via a
rectifier and DC/DC converter, which are connected in parallel to the DC bus and then recharged or
supplied to the load. In this chapter, the model of each section of the DC microgrid is built under Simulink,
and the simulation is done according to the analysis in the previous chapter.
6.2 WT system simulation
According to formulas (3-1)-(3-6) as well as a turbine power coefficient module(𝑘𝑖 , 𝑖 = 1~6), the WT
generation model is built under Simulink, as shown in Fig. 6-1(a). During the simulation process, the pitch
angle(𝛽) is set to 0∘ and the radius of the blade is 1.6𝑚, the wind speed(𝑉𝑤 ) is set from 8𝑚/𝑠 to 6𝑚/𝑠 at
2.5 𝑠𝑒𝑐𝑜𝑛𝑑𝑠, then from 6𝑚/𝑠 to 8𝑚/𝑠 at 4 seconds. It can be seen from Fig. 6-1(b) that the turbine
power coefficient (𝐶𝑝 ) oscillates greatly near the maximum power point(MPP) when the wind speed(𝑉𝑤 )
changes abruptly, but then quickly stabilizes to 𝐶𝑝_𝑀𝐴𝑋 with less oscillations after stabilization.
Fig. 6-1. (a) Simulink model of wind generator; (b) Simulation curve of wind generator.
The MPPT model of WT system is built under Simulink, as shown in Fig. 6-2, The PMSG directly calls
the model in MATLAB, the AC/DC adopts the rectifier diode, and the DC/DC converter uses the boost
type, which is utilized to vary the wind speed(𝑉𝑤 ) from 8𝑚/𝑠 to 6𝑚/𝑠 at 2.5 𝑠𝑒𝑐𝑜𝑛𝑑𝑠, then from 6𝑚/𝑠 to
8𝑚/𝑠 at 4 𝑠𝑒𝑐𝑜𝑛𝑑𝑠, and the simulation duration is set to 5 seconds. Fig. 6-3 shows the traditional P&O
Zixuan HE, 936370

45
MPPT submodule with fixed rotor speed change. The output power and rotor speed are served as input
signals to compare with the optimal values so that the perturbation direction can be detected and the MPP
can be maintained dynamically.
Fig. 6-2. Simulink model of boost converter in WT system.
Fig. 6-3. P&O MPPT controller submodule in WT system.
The MPPT simulation curve of the WT system is shown in Fig. 6-4. It can be seen that when the wind
speed changes, the fixed-step disturbance observation method oscillates greatly near the MPP. This is
mainly because the step size of traditional P&O is fixed or a wrong detecting direction when approaching
MPP. According to a research done by other scientists, it can be obviously seen that P&O with adaptive
step tracks the MPP more rapidly compared with traditional fixed-step method, as shown in Fig. 6-5
(Youssef, Mousa and Mohamed, 2020).
Zixuan HE, 936370

46
Fig. 6-4. Simulation curve of MPPT controller in WT system.
Fig. 6-5. Comparison of P&O method with fixed step and adaptive step in WT system.
6.3 PV system simulation
The PV panel module is directly called from the MATLAB library, as shown in Fig, 6-6. The
temperature is set at an ambient temperature( 25℃ ). And the irradiance variation is set to change
from 800𝑊/𝑚2 to 1000𝑊/𝑚2 at 2 seconds, then from 1000𝑊/𝑚2 to 600𝑊/𝑚2 at 3 seconds.
According to formula (11) to (15), the required parameters: 𝐼𝑠𝑐 (short-circuit current): 26.49A; 𝑉𝑜𝑐 (open-
circuit voltage): 198V, 𝐼𝑚 (current for maximum power): 24.99A, 𝑉𝑚 (voltage for maximum power): 162V.
Zixuan HE, 936370

47
Fig. 6-6. Simulink model of PV panel.
Fig. 6-7 shows the traditional P&O MPPT submodule with fixed voltage change. The voltage is
perturbed and multiplied with the current to compare with the previous output power so that the
perturbation direction can be detected and the operating power can approach the MPP in real-time. In this
article, the P&O algorithm is also used in PV module to track the optimal voltage(𝑉𝑚𝑝𝑝 ) and current(𝐼𝑚𝑝𝑝 ).
The operating voltage(𝑉) is perturbed periodically to observe the power variation so that the direction of
voltage change can be deduced to give a voltage reference (𝑉𝑟𝑒𝑓 ) (Locment, Sechilariu and Houssamo,
2010).
(a) MPPT module in PV system.
(b) MPPT algorithm submodule.
Fig. 6-7. P&O MPPT controller in PV system.
Zixuan HE, 936370

48
The whole PV system simulated consists of a PV module, a controller current source(CCM) that
converts input signals(temperature and irradiance) into an equivalent current source, a load, and a boost
converter. As shown in Fig. 5-8, the DC/DC converter of PV system is built under Simulink. The DC/DC
converter uses the boost type, which is utilized to vary the irradiance from 800𝑊/𝑚2 to 1000𝑊/𝑚2 at 2
seconds, then from 1000𝑊/𝑚2 to 600𝑊/𝑚2 at 3 seconds, and the whole simulation time is set to 5
seconds.
Fig. 6-8. Simulink model of boost converter in PV System.
It can be seen from Fig. 6-9(a) that when the irradiance input changes, there are fewer oscillations
around the MPP in P&O MPPT controller. Moreover, the larger the changing amplitude is, the larger the
overshoot around the MPP is. And PV system has more oscillations when tracking the MPP provided the
same conditions as the WT system. Since the superiority of adaptive step size over fixed step is
demonstrated in WT system, this part makes a comparison between P&O and INC methods in PV system.
(a) Simulation curve of P&O MPPT controller.
Zixuan HE, 936370

49
(b) Simulation curve of INC MPPT controller.
Fig. 6-9. Simulation curve of MPPT controller in PV system.
It can be noticed that the INC method has a quicker response time than INC method at the initial stage.
As shown in Fig. 6-9(b), the tracking process of INC method experiences higher overshoots and fiercer
oscillations when the irradiance varies abruptly. However, when the irradiance is unchanged, the MPP
controlled by INC algorithm is more stable than that of P&O method.
The main reason may be INC controller controls two changing values(voltage and current)
simultaneously in order for the voltage operating point where the instantaneous conductance (I/V) is equal
to the incremental conductance (dI/dV). As for P&O method, operating voltage is perturbed first, then the
direction is detected, after a step change of voltage, current value is adjusted to match with the MPP.
Although the control logic of P&O method is simple and orderly, it takes more time to track the MPP.
6.4 ESS module simulation
As shown in Fig. 6-10, the DC/DC converter of ESS module is built under Simulink. This module
consists of a LIB sector that is directly called from the MATLAB library, a bidirectional DC/DC converter
utilized to switch the circuit to different modes depending on the SoC and load demand, with an objective
to regulate the DC bus voltage. The parameters of the LIB sector are Nominal Voltage: 200𝑉; Rated
Capacity: 9.6𝐴ℎ; Initial SoC: 80%; Battery Response Time: 30𝑠.
Zixuan HE, 936370

50
Fig. 6-10. Simulink model of bidirectional converter in ESS System.

∗
As shown in Fig. 6-11, the DC voltage of the DC bus(𝑉𝑑𝑐 ) and the voltage reference(𝑉𝑑𝑐 : 400𝑉) are set
as the inputs of the ADRC controller to generate the battery current reference, which is used as an input
signal for the battery control module. The parameters for the ADRC controller are 𝑟 = 100, 𝑤0 = 34, 𝑏 =
69, and 𝑘 = 60. Then, the output current reference is sent to the battery control module, with an objective
to compare with the operating battery current so that the optimal current and voltage are updated and
delivered to the DC/DC converter.
Fig. 6-11. Simulink model of control part in ESS module.
As shown in Fig. 6-12, the simulation curve of LIB sector is demonstrated. During the charging
period (0~1.5𝑠), the battery voltage gradually decreases, and the SoC gradually increases. Conversely,
during discharging mode (1.5~5𝑠), there is a decrement in battery voltage and an increment in SoC,
and the curve gradient is dependent on the operating conditions of WT and PV systems. Moreover, the
charge stored in the capacitor keeps accumulating. If an ultracapacitor is integrated into the ESS module,
those charges can also be used to supply the load.
Zixuan HE, 936370

51
Fig. 6-12. Simulation curve of LIB sector.
6.5 DC voltage simulation
As shown in Fig. 6-13, the EM strategy of the LIB sector is formulated under Simulink. The current
reference battery( 𝑖 𝑏𝑎𝑡𝑟𝑒𝑓 ) given out by the ADRC module is utilized as criteria to judge which mode to
undergo, and the feedback input refers to the battery current in the last scanning cycle.
Fig. 6-13. Simulink model of EM strategy.
Fig. 6-14 reflects the charging and discharging conditions of the LIB when the external conditions
(irradiance and wind speed) change. It can be concluded that: between 0 and 1.5s, the output power of the
Zixuan HE, 936370

52
power generation system is not sufficient to meet the demand of the load, so it works in mode 1(𝐾 = 1),
and the battery participates in the regulation to keep the bus voltage stable. Between 1.5 and 5s, the power
output of the power generation system is sufficient to meet the load demand and there is still excess power,
so it is working in mode 1 (𝐾 = 0), therefore charging the battery, the battery output is negative.
Furthermore, a comparison is made between the ADRC scheme and traditional PI controller, it can be
noticed that ADRC takes less response time while more oscillations occur when tracking the DC
voltage(400V) and climate changes.
(a) ADRC-based DC voltage control.
(b) PI-based DC voltage control.
Fig. 6-14. Simulation curve of DC bus voltage.
Zixuan HE, 936370

53
6.6 Summary
In nutshell, this chapter presents the Simulink model and simulation curve of each DC microgrid
section and analyzes the results using different control groups. In WT system modelling, a simulation
result that compares P&O method with fixed and adaptive step size is cited to show the superiority of the
latter. In PV system, an INC MPPT controller is also programmed using MATLAB to make a comparison
with P&O algorithm to weigh the pros and cons. And in ESS module, the ADRC-based DC voltage
control scheme is simulated in contrast to the traditional PI-based approach to confirm the advantage of
ADRC to reduce response time and oscillations around MPP when environment condition changes. And
an overview of the whole standalone DC MG Simulink model is attached, as depicted in Fig. 6-15.
Fig. 6-15. Simulink model of proposed standalone DC microgrid.
Zixuan HE, 936370

54
Chapter 7 Conclusion and Outlook
7.1 Conclusion
This article focuses on the following issues: firstly, to understand how DC MGs works, including the
working principles of wind generators, solar panels, and LIBs based on mathematical deviation; secondly,
to know how DC/DC converter works with MPPT algorithms track the MPP; thirdly, to study how to use
ADRC optimize the DC voltage regulation in the ESS module. The main contents of this article focus on:
(1) A Simulink model of standalone DC microgrid is built based on the analysis of wind energy
system, solar energy system, and energy storage system. And the correctness of the model is
verified through simulation.
(2) Based on the existing algorithm for MPPT, the simulation is carried out to verify the feasibility of
the algorithm by improving the P&O and INC methods with a boost type DC/DC power. And the
comparison is discussed between P&O algorithm and INC algorithm in PV system.
(3) Based on the energy management strategy, a bidirectional DC/DC converter is utilized in ESS
module to switch the LIB sector to different working conditions. An ADRC algorithm is used to
optimize the DC voltage regulation compared with traditional PI controller.
7.2 Further research objective
There are still some shortcomings in this paper, and the following aspects need to be improved in
subsequent research:
(1) Further study of battery charging and discharging strategies such as Battery-UC based HESS
modules can be used to extend the lifespan of LIBs by utilizing ultracapacitor to provide transient
power during high-frequency power exchange to improve its life span and EM accuracy.
(2) This paper only considers the standalone(off-grid) conditions of the DC microgrid system. In the
future research process, the grid-connected model can be further considered to improve the
applicability of the MG systems.
Zixuan HE, 936370

55
(3) The factors considered for the simulation of the RES generation systems are not yet complete,
such as the influence of temperature variation on solar cells and the effect of air density on the
output of wind turbines.
(4) The improved MPPT algorithms that use adaptive step change to track MPP intelligently are not
applied in this article, the future study could focus on how to improve the MPPT algorithms for
DG systems.
Reference
1. Abdullah, M.A. et al. (2012) “A review of maximum power point tracking algorithms for wind energy
systems,” Renewable and Sustainable Energy Reviews, pp. 3220–3227. doi:10.1016/j.rser.2012.02.016.
2. Amiri, H. et al. (2020) “Fully decentralized robust backstepping voltage control of photovoltaic
systems for DC islanded microgrids based on disturbance observer method,” ISA Transactions, 101,
pp. 471–481. doi:10.1016/j.isatra.2020.02.006.
3. Armghan, H. et al. (2020) “Nonlinear integral backstepping based control of a DC microgrid with
renewable generation and energy storage systems,” International Journal of Electrical Power and
Energy Systems, 117(July 2019), p. 105613. doi:10.1016/j.ijepes.2019.105613.
4. Atia, R. and Yamada, N. (2016) “Sizing and Analysis of Renewable Energy and Battery Systems in
Residential Microgrids,” IEEE Transactions on Smart Grid, 7(3), pp. 1204–1213.
doi:10.1109/TSG.2016.2519541.
5. B. S., H., Yan, B. and Rudy, S. (2013) “Review of Microgrid Technology,” in 2013 International
Conference on QiR. IEEE. Available at:
https://ieeexplore.ieee.org/abstract/document/6632550/authors#authors (Accessed: February 28, 2022).
6. Badwawi, R. al, Abusara, M. and Mallick, T. (2015) “A Review of Hybrid Solar PV and Wind Energy
System,” Smart Science, 3(3), pp. 127–138. doi:10.1080/23080477.2015.11665647.
7. Bari, A. et al. (2014) “Challenges in the smart grid applications: An overview,” International Journal
of Distributed Sensor Networks, 2014. doi:10.1155/2014/974682.
Zixuan HE, 936370

56
8. Baroudi, J.A., Dinavahi, V. and Knight, A.M. (2007) “A review of power converter topologies for
wind generators,” Renewable Energy, 32(14), pp. 2369–2385. doi:10.1016/j.renene.2006.12.002.
9. Belfkira, R., Zhang, L. and Barakat, G. (2011) “Optimal sizing study of hybrid wind/PV/diesel power
generation unit,” Solar Energy, 85(1), pp. 100–110. doi:10.1016/j.solener.2010.10.018.
10. Bianchi, F.D., de Battista, H. and Mantz, R.J. (2007) Wind turbine control systems : principles,
modelling and gain scheduling design. Springer.
11. Buehring, I.K. and Freris, L.L. (1981) “CONTROL POLICIES FOR WIND-ENERGY
CONVERSION SYSTEMS.,” IEE Proceedings C: Generation Transmission and Distribution, 128(5),
pp. 253–261. doi:10.1049/IP-C.1981.0043/CITE/REFWORKS.
12. Carrillo, C. et al. (2013) “Review of power curve modelling for wind turbines,” Renewable and
Sustainable Energy Reviews, 21, pp. 572–581. doi:10.1016/J.RSER.2013.01.012.
13. Celik, B. et al. (2017) “Electric energy management in residential areas through coordination of
multiple smart homes,” Renewable and Sustainable Energy Reviews, 80, pp. 260–275.
doi:10.1016/J.RSER.2017.05.118.
14. Chen, F. et al. (2015) “A nonlinear droop method to improve voltage regulation and load sharing in
DC systems,” 2015 IEEE 1st International Conference on Direct Current Microgrids, ICDCM 2015,
pp. 45–50. doi:10.1109/ICDCM.2015.7152008.
15. Coelho, R.F., Concer, F.M. and Martins, D.C. (2010) “A MPPT approach based on temperature
measurements applied in PV systems,” 2010 IEEE International Conference on Sustainable Energy
Technologies, ICSET 2010 [Preprint]. doi:10.1109/ICSET.2010.5684440.
16. Cucuzzella, M. et al. (2018) “Sliding mode voltage control of boost converters in DC microgrids,”
Control Engineering Practice, 73, pp. 161–170. doi:10.1016/j.conengprac.2018.01.009.
17. Dragicevic, T. et al. (2014) “Supervisory control of an adaptive-droop regulated DC microgrid with
battery management capability,” IEEE Transactions on Power Electronics, 29(2), pp. 695–706.
doi:10.1109/TPEL.2013.2257857.
18. Elmouatamid, A. et al. (2020) “Review of Control and Energy Management Approaches in Micro-
Grid Systems,” Energies 2021, Vol. 14, Page 168, 14(1), p. 168. doi:10.3390/EN14010168.
Zixuan HE, 936370

57
19. Elmouatamid, A. et al. (2021) “A MicroGrid System Infrastructure Implementing IoT/Big-Data
Technologies for Efficient Energy Management in Buildings,” Green Energy and Technology, pp.
571–600. doi:10.1007/978-3-030-64565-6_20.
20. Espín‐Sarzosa, D., Palma-Behnke, R. and Núñez‐Mata, O. (2020) “Energy Management Systems for
Microgrids: Main Existing Trends in Centralized Control Architectures,” Energies 2020, Vol. 13, Page
547, 13(3), p. 547. doi:10.3390/EN13030547.
21. Esram, T. and Chapman, P.L. (2007) “Comparison of photovoltaic array maximum power point
tracking techniques,” IEEE Transactions on Energy Conversion, 22(2), pp. 439–449.
doi:10.1109/TEC.2006.874230.
22. Farhani, S., Barhoumi, E.M. and Bacha, F. (2021) “Design and hardware investigation of a new
configuration of an isolated DC-DC converter for fuel cell vehicle,” Ain Shams Engineering Journal,
12(1), pp. 591–598. doi:10.1016/j.asej.2020.07.014.
23. Femia, N. et al. (2013) “Power Electronics and Control Techniques for Maximum Energy Harvesting
in Photovoltaic Systems,” IEEE Industrial Electronics Magazine, 7(3), pp. 66–67.
doi:10.1109/mie.2013.2272239.
24. Gao, F. et al. (2019) “Primary and secondary control in DC microgrids: a review,” Journal of Modern
Power Systems and Clean Energy, 7(2), pp. 227–242. doi:10.1007/S40565-018-0466-5.
25. González-Romera, E. et al. (2020) “Secondary Control for Storage Power Converters in Isolated
Nanogrids to Allow Peer-to-Peer Power Sharing,” Electronics 2020, Vol. 9, Page 140, 9(1), p. 140.
doi:10.3390/ELECTRONICS9010140.
26. Guerrero, J.M. et al. (2013) “Advanced control architectures for intelligent microgridspart i:
Decentralized and hierarchical control,” IEEE Transactions on Industrial Electronics, 60(4), pp.
1254–1262. doi:10.1109/TIE.2012.2194969.
27. Hossain, M.A. et al. (2019) “Evolution of microgrids with converter-interfaced generations:
Challenges and opportunities,” International Journal of Electrical Power and Energy Systems.
Elsevier Ltd, pp. 160–186. doi:10.1016/j.ijepes.2019.01.038.
28. Hua, H. et al. (2018) “Voltage control for uncertain stochastic nonlinear system with application to
Zixuan HE, 936370

58
energy Internet: Non-fragile robust H∞ approach,” Journal of Mathematical Analysis and
Applications, 463(1), pp. 93–110. doi:10.1016/j.jmaa.2018.03.002.
29. Ishaque, K., Salam, Z. and Syafaruddin (2011) “A comprehensive MATLAB Simulink PV system
simulator with partial shading capability based on two-diode model,” Solar Energy, 85(9), pp. 2217–
2227. doi:10.1016/J.SOLENER.2011.06.008.
30. Jinbo, M. et al. (2014) “Fixed and adaptive Step HCC algorithms for MPPT of the cylinders of
magnus wind turbines,” IET Conference Publications, 2014(CP651). doi:10.1049/CP.2014.0921.
31. Jing, W. et al. (2016) Battery-Supercapacitor Hybrid Energy Storage System in Standalone DC
Microgrids: A Review, IET Renewable Power Generation.
32. Kakosimos, P.E. and Kladas, A.G. (2011) “Implementation of photovoltaic array MPPT through fixed
step predictive control technique,” Renewable Energy, 36(9), pp. 2508–2514.
doi:10.1016/J.RENENE.2011.02.021.
33. Kim, J.W., Choi, H.S. and Cho, B.H. (2002) “A novel droop method for converter parallel operation,”
IEEE Transactions on Power Electronics, 17(1), pp. 25–32. doi:10.1109/63.988666.
34. Koutroulis, E. and Kalaitzakis, K. (2006) “Design of a maximum power tracking system for wind-
energy-conversion applications,” IEEE Transactions on Industrial Electronics, 53(2), pp. 486–494.
doi:10.1109/TIE.2006.870658.
35. Kumar, D. and Chatterjee, K. (2016) “A review of conventional and advanced MPPT algorithms for
wind energy systems,” Renewable and Sustainable Energy Reviews, 55, pp. 957–970.
doi:10.1016/j.rser.2015.11.013.
36. Lakshmi, M. and Hemamalini, S. (2017) “Nonisolated high gain DC-DC converter for DC microgrids,”
IEEE Transactions on Industrial Electronics, 65(2), pp. 1205–1212. doi:10.1109/TIE.2017.2733463.
37. Learn about electronics (2020a) Power Supplies: Boost converter. Available at:
https://www.learnabout-electronics.org/PSU/psu32.php (Accessed: November 27, 2021).
38. Learn about electronics (2020b) Power supplies: Buck converter, Learn about electronics. Available at:
https://www.learnabout-electronics.org/PSU/psu31.php (Accessed: November 27, 2021).
39. Levi, E., Gofer, Y. and Aurbach, D. (2010) “On the way to rechargeable Mg batteries: The challenge
Zixuan HE, 936370

59
of new cathode materials,” Chemistry of Materials, 22(3), pp. 860–868. doi:10.1021/cm9016497.
40. Li, J. (2018) Hydrogen Energy Country Overview: China. Available at:
https://www.energy.gov/sites/prod/files/2018/10/f56/fcto-infrastructure-workshop-2018-4-li.pdf.
41. LI, T., ZHU, J. and LIU, Z. (2012) “Speed observation of asynchronous motor based on improved
ESO,” Harbin University of Science and Technology, 17(1), p. 50.
42. Lin, R., Zhao, Y. and Wu, B. (2020) “Toward a hydrogen society : Hydrogen and smart grid
integration,” International Journal of Hydrogen Energy, 45(39), pp. 20164–20175.
doi:10.1016/j.ijhydene.2020.01.047.
43. Liu, B. et al. (2007) “Analysis and improvement of maximum power point tracking algorithm based
on incremental conductance method for photovoltaic array,” Proceedings of the International
Conference on Power Electronics and Drive Systems, pp. 637–641. doi:10.1109/PEDS.2007.4487768.
44. Liu, Fangrui et al. (2008) “A variable step size INC MPPT method for PV systems,” IEEE
Transactions on Industrial Electronics, 55(7), pp. 2622–2628. doi:10.1109/TIE.2008.920550.
45. Locment, F., Sechilariu, M. and Houssamo, I. (2010) “Energy efficiency experimental tests
comparison of P&O algorithm for PV power system,” Proceedings of EPE-PEMC 2010 - 14th
International Power Electronics and Motion Control Conference [Preprint].
doi:10.1109/EPEPEMC.2010.5606774.
46. Lu, H. and Fu, Z. (2011) “Feasibility analysis of solar-wind complementary power generation system
applied in engineering,” SHANXI ARCHITECTURE, 37(10), pp. 126–127. doi:10.13719/j.cnki.cn14-
1279/tu.2011.10.154.
47. Lucas-Marcillo, K.E. et al. (2019) “Novel robust methodology for controller design aiming to ensure
DC microgrid stability under CPL power variation,” IEEE Access, 7, pp. 64206–64222.
doi:10.1109/ACCESS.2019.2915027.
48. Luna, A.C. et al. (2017) “Mixed-integer-linear-programming-based energy management system for
hybrid PV-wind-battery microgrids: Modeling, design, and experimental verification,” IEEE
Transactions on Power Electronics, 32(4), pp. 2769–2783. doi:10.1109/TPEL.2016.2581021.
49. Lynser, F.B. et al. (2018) “Comparative Analysis of Different Control Schemes for DC-DC Converter:
Zixuan HE, 936370

60
A Review,” ADBU Journal of Electrical and Electronics Engineering (AJEEE), 2. Available at:
www.tinyurl.com/ajeee-adbu.
50. Mahmud, M.A. et al. (2019) “Robust Nonlinear Adaptive Feedback Linearizing Decentralized
Controller Design for Islanded DC Microgrids,” in IEEE Transactions on Industry Applications.
Institute of Electrical and Electronics Engineers Inc., pp. 5343–5352. doi:10.1109/TIA.2019.2921028.
51. Mark, H. (2010) Stand-alone Solar Electric Systems: The Earthscan Expert Handbook for Planning,
Design and Installation. 1st edn. London: Routledge. doi:https://doi.org/10.4324/9781849776509.
52. Mehdi, M. et al. (2020) “Robust control of a DC microgrid under parametric uncertainty and
disturbances,” Electric Power Systems Research, 179. doi:10.1016/j.epsr.2019.106074.
53. Meraj, S.T. et al. (2021) “A hybrid T-type (HT-type) multilevel inverter with reduced components,”
Ain Shams Engineering Journal, 12(2), pp. 1959–1971. doi:10.1016/j.asej.2020.12.010.
54. Messai, A. et al. (2011) “Maximum power point tracking using a GA optimized fuzzy logic controller
and its FPGA implementation,” Solar Energy, 85(2), pp. 265–277.
doi:10.1016/J.SOLENER.2010.12.004.
55. Molzahn, D.K. et al. (2017) “A Survey of Distributed Optimization and Control Algorithms for
Electric Power Systems,” IEEE Transactions on Smart Grid, 8(6), pp. 2941–2962.
doi:10.1109/TSG.2017.2720471.
56. Mosa, M. et al. (2012) “Modified MPPT with using model predictive control for multilevel boost
converter,” IECON Proceedings (Industrial Electronics Conference), pp. 5080–5085.
doi:10.1109/IECON.2012.6389559.
57. Nelson, D.B., Nehrir, M.H. and Wang, C. (2006) “Unit sizing and cost analysis of stand-alone hybrid
wind/PV/fuel cell power generation systems,” Renewable Energy, 31(10), pp. 1641–1656.
doi:10.1016/j.renene.2005.08.031.
58. Nema, P., Nema, R.K. and Rangnekar, S. (2009) “A current and future state of art development of
hybrid energy system using wind and PV-solar: A review,” Renewable and Sustainable Energy
Reviews, pp. 2096–2103. doi:10.1016/j.rser.2008.10.006.
59. Padmanaban, S. et al. (2018) “An original transformer and switched-capacitor (T & SC)-based
Zixuan HE, 936370

61
extension for DC-DC boost converter for high-voltage/low-current renewable energy applications:
Hardware implementation of a new T & SC boost converter,” Energies, 11(4), p. 783.
doi:10.3390/en11040783.
60. Peng, W., Feng, L. and Song, Y. (2013) “A Novel Maximum Power Point Tracking Control Method in
Wind Turbine Application,” in Proceedings of the Chinese Control Conference.
61. Prabaharan, N. et al. (2018) “An overview of control techniques and technical challenge for inverters
in micro grid,” Hybrid-renewable energy systems in microgrids: Integration, developments and
control, pp. 97–107. doi:10.1016/B978-0-08-102493-5.00006-6.
62. Prabhakaran, P., Goyal, Y. and Agarwal, V. (2018) “Novel nonlinear droop control techniques to
overcome the load sharing and voltage regulation issues in DC Microgrid,” IEEE Transactions on
Power Electronics, 33(5), pp. 4477–4487. doi:10.1109/TPEL.2017.2723045.
63. Pradeep, A. et al. (2012) “Comparison of MPPT Algorithms for DC-DC Converters Based PV
Systems,” International Journal of Advanced Research in Electrical, Electronics and Instrumentation
Engineering, 1(1). Available at: www.ijareeie.com (Accessed: November 26, 2021).
64. Ragheb, M. (2014) “WIND ENERGY CONVERSION THEORY, BETZ EQUATION,” Wind Energie
[Preprint].
65. Sabir, M., Abd Ghani, H. and Abd Elhamid, L. (2015) “A New Neural Networks MPPT controller for
PV Systems,” in 2015 6th International Renewable Energy Congress (IREC).
66. Sivakumar, S. et al. (2016) “An assessment on performance of DC-DC converters for renewable
energy applications,” Renewable and Sustainable Energy Reviews. Elsevier, pp. 1475–1485.
doi:10.1016/j.rser.2015.12.057.
67. Smith, K. et al. (2017) “Life prediction model for grid-connected Li-ion battery energy storage
system,” in Proceedings of the American Control Conference. Institute of Electrical and Electronics
Engineers Inc., pp. 4062–4068. doi:10.23919/ACC.2017.7963578.
68. Thongam, J.S. and Mohand, O. (2011) “Fundamental and Advanced Topics in Wind Power,” in Dr.
Rupp Carriveau (ed.) Fundamental and Advanced Topics in Wind Power, pp. 342–347. Available at:
https://sc.panda321.com/extdomains/books.google.com/books?hl=zh-
Zixuan HE, 936370

62
CN&lr=&id=P9CPDwAAQBAJ&oi=fnd&pg=PA339&dq=9%5D+Thongam+JS,+Ouhrouche+M.+M
PPT+control+methods+in+wind+energy+conver%3Fsion+systems.+In:+Carriveau+R,+editor.+Adv.+
Top.+Wind+Power,+InTech%3B+2011.+p.+339%E2%80%9360.&ots=iDR1de3DgY&sig=PzIiU3Hi
B6Q4IiTsueQ2WIEj0M0#v=onepage&q&f=false (Accessed: November 25, 2021).
69. Ullah, S. et al. (2020) “The current state of Distributed Renewable Generation, challenges of
interconnection and opportunities for energy conversion based DC microgrids,” Journal of Cleaner
Production. Elsevier Ltd. doi:10.1016/j.jclepro.2020.122777.
70. Ustun, T.S. and Mekhilef, S. (2010) “Effects of a Static Synchronous Series Compensator (SSSC)
Based on a Soft Switching 48-Pulse PWM Inverter on the Power Demand from the Grid,” Journal of
Power Electronics, 10(1), pp. 85–90. doi:10.6113/JPE.2010.10.1.085.
71. Van, N.D. et al. (2020) “A Hierarchical Control System for Autonomous Driving towards Urban
Challenges,” Applied Sciences 2020, Vol. 10, Page 3543, 10(10), p. 3543. doi:10.3390/APP10103543.
72. Wang, Q. and Chang, L. (2004) “An intelligent maximum power extraction algorithm for inverter-
based variable speed wind turbine systems,” IEEE Transactions on Power Electronics, 19(5), pp.
1242–1249. doi:10.1109/TPEL.2004.833459.
73. Warnier, M. et al. (2017) “Distributed monitoring for the prevention of cascading failures in
operational power grids,” International Journal of Critical Infrastructure Protection, 17, pp. 15–27.
doi:10.1016/J.IJCIP.2017.03.003.
74. Xu, G., Sha, D. and Liao, X. (2015) “Decentralized inverse-droop control for input-series-output-
parallel DC-DC Converters,” IEEE Transactions on Power Electronics, 30(9), pp. 4621–4625.
doi:10.1109/TPEL.2015.2396898.
75. Yamashita, D.Y., Vechiu, I. and Gaubert, J.P. (2020) “A review of hierarchical control for building
microgrids,” Renewable and Sustainable Energy Reviews, 118, p. 109523.
doi:10.1016/J.RSER.2019.109523.
76. Yang, H., Lu, L. and Zhou, W. (2007) “A novel optimization sizing model for hybrid solar-wind
power generation system,” Solar Energy, 81(1), pp. 76–84. doi:10.1016/j.solener.2006.06.010.
77. Youssef, A.R., Mousa, H.H.H. and Mohamed, E.E.M. (2020) “Development of self-adaptive P&O
Zixuan HE, 936370

63
MPPT algorithm for wind generation systems with concentrated search area,” Renewable Energy, 154,
pp. 875–893. doi:10.1016/j.renene.2020.03.050.
78. Zanganeh, K.E. and Shafeen, A. (2007) “A novel process integration , optimization and design
approach for large-scale implementation of oxy-fired coal power plants with CO 2 capture,”
international journal of greenhouse gas control, 1(1), pp. 47–54. doi:10.1016/S1750-5836(07)00035-
7.
79. Zhang, X. and Zhou, M. (2020) “ADRC-BASED ROBUST DC VOLTAGE CONTROL OF THE
GRID-CONNECTED VSC IN DC MICROGRID,” U.P.B. Sci. Bull., Series C, 82(2), p. 2020.
80. Zhou, K. et al. (2018) “Active disturbance rejection control of PMSM speed control system,” Electric
Machines and Control, 22(2), pp. 57–63.
Zixuan HE, 936370

64
Appendix
The code of INC MPPT algorithm in PV system:
Zixuan HE, 936370

65
Acknowledgement
First and foremost, I have to thank my research supervisor, Dr. Karam Eliker. His guidance and advice
carried me through all the stages of writing my project. Every time I asked him questions on my research
or academic writing, he always gave me helpful references and prompt suggestions to make my research
and report more convincing and professional.
I would also like to thank my colleges: Liverpool John Moores University(LJMU) and University of
Shanghai for Science and Technology(USST). Although many research work can only be done online
during the COVID-19 pandemic period, the staff have been always trying their best to offer students a
comfortable research environment.
Finally, none of this could have happened without my family. My mother offer her continuous support
and encouragement throughout my study career. Without her dedication to my education, I can never be a
person who take everything seriously and treasure every opportunity.
Zixuan HE, 936370

66

Final Report - Zixuan HE

Uploaded by

Copyright:

Available Formats

Final Report - Zixuan HE

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Final Report - Zixuan HE

Uploaded by

Copyright:

Available Formats

6101SBC:

Title: An ADRC Controlled

LJMU number: 936370

Advisor: Dr. Karam Eliker

Zixuan HE, 936370

Zixuan HE, 936370

Figure No. Figure Title Page No.

Zixuan HE, 936370

Table No. Table Title Page No.

Zixuan HE, 936370

Abbreviation Full Name

Zixuan HE, 936370

1.1 Research background

regions(Ullah et al., 2020).

Zixuan HE, 936370

1.2 Problem statement

this research are proposed as:

Zixuan HE, 936370

(1) The power converter modelling for DC microgrid in standalone mode

ADRC controller to control the LIB sector

1.3 Research objective

Fig. 1-1. The considered DC microgrid.

The following are the specific experimental objectives to be done:

Zixuan HE, 936370

system to charging mode or discharging mode.

1.4 Research contribution

achievements of this DC microgrid research are listed as followed:

scheme is compared with PI controller to regulate the DC voltage more intelligently.

1.5 Thesis structure

chapter 3, the mathematical modelling of DC microgrid components, including WT system, PV system

Zixuan HE, 936370

Fig. 1-2. Thesis structure of this article.

Zixuan HE, 936370

Chapter 2 Literature Review

2.2 Standalone MG system

Zixuan HE, 936370

No. Challenges Solutions

Zixuan HE, 936370

2.3 Energy management strategy

Fig. 2-2. The hierarchical structure of an MG control system.

strategies: Centralized control, decentralized control, and hierarchal control.

2.3.1 Centralized control

Centralized control scheme uses a central controller(CC), which is characterized by a high-performance

control(Warnier et al., 2017).

Fig. 2-3. The layout of centralized control approaches.

Zixuan HE, 936370

their limitations, especially when dealing with large-scale hybrid systems.

2.3.2 Decentralized control

communication networks(Yamashita, Vechiu and Gaubert, 2020).

Fig. 2-4. The layout of decentralized control approaches.

According to the availability of communication networks, decentralized control can be subclassified

Zixuan HE, 936370

during MG system installation(Espín‐Sarzosa, Palma-Behnke and Núñez‐Mata, 2020).

2.3.3 Hierarchical control

Fig. 2-5. The layout of hierarchical control approaches.

Zixuan HE, 936370

modes(Guerrero et al., 2013).

2.3.4 A comparison of EM strategy

Table 2-2. A comparison of EM strategies.