Frequency Constrained Optimal Power Flow

Incorporating UPFC Controller Using Driving

Training-Based Optimization
Adhit Roy
Dr. B. C. Roy Engineering College
Susanta Dutta
Dr. B. C. Roy Engineering College
Anagha Bhattacharya
NIT
Soumen Biswas
Dr. B. C. Roy Engineering College
Rajesh kumar chatterjee
Usha Martin University
Md Amir Khusru Akhtar
Usha Martin University
Mohit Kumar
MIT Art, Design and Technology University
Sahil Verma
Universidade Federal do Piauí
Ruba Abu Khurma
Middle East University
Saurav Mallik
Harvard T H Chan School of Public Health
Mohd Asif Shah

Kardan University

Article

Keywords: Optimal Power Flow (OPF), Load Frequency Control (LFC), Frequency based Security
Constraint, Uni ed power ow controller (UPFC), Driving training-based optimization(DTBO)

Posted Date: June 11th, 2024

DOI: https://doi.org/10.21203/rs.3.rs-4493212/v1

License:   This work is licensed under a Creative Commons Attribution 4.0 International License.
Read Full License

Additional Declarations: No competing interests reported.

Frequency Constrained Optimal Power Flow Incorporating UPFC
Controller Using Driving Training-Based Optimization
Adhit Roya , Susanta Duttaa , Anagha Bhattacharyab , Soumen Biswas*a , Rajesh kumar chatterjeec ,
Md Amir Khusru Akhtar c , Mohit Kumare , Sahil Verma f,g , Ruba Abu Khurmah,i , Saurav Mallik*j,k
and Mohd Asif Shah*l,∗∗
a Department of Electrical Engineering, Dr. B. C. Roy Engineering College, Durgapur, India; [email protected], [email protected],
[email protected]
b Department of Electrical Engineering, NIT,Mizoram,India; [email protected]
c Department of Computing & IT Department, Usha Martin University, India; [email protected]; [email protected]
e Department of IT, MIT Art, Design and Technology University,India
f Universidade Federal do Piauí, Teresina, Piauí, 64049-550, Brazil (e-mail: [email protected])
g Manipal University, Rajasthan, 303007, India; [email protected]
h MEU Research Unit, Faculty of Information Technology, Middle East University, Amman, 11831, Jordan (e-mail: [email protected])
i Applied Science Research Center, Applied Science Private University, Amman 11931, Jordan
j Department of Environmental Health, Harvard T H Chan School of Public Health, Boston, MA 02115, USA (e-mail: [email protected],

[email protected])
k Department of Pharmacology & Toxicology, University of Arizona, Tucson, MA 85721, USA
l Department of Economics, Kardan University, Parwane Du, 1001, Kabul, Afghanistan (e-mail: [email protected])
∗ Corresponding authors’ e-mail: [email protected], [email protected], [email protected])

ARTICLE INFO ABSTRACT

Keywords: Optimal power flow (OPF) solution is being attained for IEEE 30-bus system considering thermal
Optimal Power Flow (OPF) power (TP) sources. This paper presents a methodology to incorporate UPFC with frequency-based
Load Frequency Control (LFC) security constraints in the OPF problem framework. Objectives include reducing generation costs,
Frequency based Security Constraint emissions, and frequency deviation. This paper has looked at three different scenarios: the best
Unified power flow controller (UPFC) power flow when there is no frequency-based security limitation, the best power flow when there
Driving training-based optimization is a frequency-based security constraint, and the combination of UPFC and frequency-based security
(DTBO) constraint. It is necessary to maintain the system frequency within a safe range in order to display the
balance between generation and consumption. Therefore, the power flow optimization solution should
provide frequency stability in addition to provide the lowest possible generation cost under operational
conditions. To achieve this solution, a new constraint is implemented on the power dispatch problem
to achieve frequency stability. The inclusion of this limitation is demonstrated to induce the OPF to
choose greater inertia generators more favorably when it comes to meet the frequency based security
criterion. Outcomes of the test reveal a better result in resolving the OPF problem, by incorporating
a unified power flow controller (UPFC) with frequency-based security constraint OPF problem. A
driving training-based optimization (DTBO) approach has been employed to obtain optimal solutions
without violating any equality and inequality network constraints.

1. Introduction lem aims to identify the optimal operational state of a power

source to meet the load side demand while accounting for se-
The optimal power flow (OPF) is one of the most im-
curity constraints. Since inertia constants are becoming less
portant method for power systems management and con-
of a significant concern, frequency deviation will increase at
trol. The primary objectives of OPF are to minimise power lower loads. The system needs to have sufficient spinning
loss, minimise generating costs, preserve voltage stability reserve and inertia to keep frequency within safe bounds.
and emissions, all while keeping multiple system restrictions However, due to the lack of inertia and regulation capac-
at their ideal settings. During the optimisation process, the ity brought about by the large degree of generating variety
system operates at its best when the voltage of the buses,
in contemporary power networks, frequency based security
the intricate power distribution in the lines, and the sched-
now faces additional difficulties [1, 2]. Therefore, the im-
uled generator power are all in line with the necessary oper- pact of frequency based security must be included in the op-
ational condition of the system. It is necessary to pay close timal power flow (OPF) solution. Due to its limited ability to
attention to the power generator’s capacities, the generator regulate frequency, the least cost dispatch system may nev-
bus voltage, and the power flow in the line. The OPF prob- ertheless violate frequency based security even though it of-
ORCID (s): 0000-0002-3433-5808 (A. Bhattacharya); ten satisfies standard operating conditions. In its most basic
0000-0002-3433-5808 (S. Biswas*); 0000-0002-3433-5808 (R.k. chatterjee); form, OPF aims to maximise social welfare or minimise pro-
0000-0002-3433-5808 (M.A.K.A. ); 0000-0002-3433-5808 (M. Kumar);
duction costs while taking into account equality constraints
0000-0003-3136-4029 (S.V. ); 0000-0002-8234-9374 (R.A. Khurma);
0000-0003-4107-6784 (S. Mallik*); 0009-0000-2821-5423 (M.A. Shah*) that describe the balance between real and reactive power

𝐶𝑜𝑟𝑟𝑒𝑠𝑝𝑜𝑛𝑑𝑖𝑛𝑔 𝑎𝑢𝑡ℎ𝑜𝑟 𝑚𝑎𝑖𝑙 𝑖𝑑: [email protected] Page 1 of 17

 Frequency Constrained OPF Incorporating UPFC Controller Using DTBO

in addition to the voltage magnitude limitations, equipment The lack of quantification of the impact of greater pene-
capability limits, and other operating parameters. In order tration of less inertia at lower load, the grid frequency is one
to ensure that a system’s frequency deviation stays within a of the main weaknesses of the research that have been de-
safe range under typical circumstances, the OPF will take scribed in the literature. Furthermore, the literature that has
the frequency stability restriction into consideration. already been published does not yet have a single index that
Researchers have presented a few conventional and measures the frequency stability of a system. Additionally,
heuristic optimisation strategies over the past few decades to given the tendency towards a decline in grid inertia, it’s crit-
solve the OPF challenges. Traditional techniques encompass ical to assess how sensitive the electrical grid’s frequency
the gradient approach, interior point and newton formula- stability is to variations in inertia. In this research, a control
tion, as well as linear, nonlinear, and quadratic programming strategy with minimal transient deviations and zero steady
[3]. In order to enhance convergence time, Kohli applied the state error is proposed to maintain power supply quality and
chaotic grey wolf optimisation algorithm [4]. To improve dependability. It becomes challenging to stabilise frequency
the solution, A. A. Heidari was introduced an efficient grey fluctuations in an interconnected power system when it is
wolf optimizer with Lévy flight (LGWO) [5]. A new tech- used in a future competitive economy. This article extends
nique called the ant lion optimisation (ALO) algorithm with the OPF framework to optimise the problem with a complete
the Lévy flight operator [6], suggested by Mirjalili. A slime set of system operating constraints, including the frequency
mould algorithm is presented by Khunkitti et al. [7] to solve based security constraint. In this case, OPF not only reduces
multi-objective optimal power flow issues. Crowd search al- production costs but also assigns greater inertia producers
gorithms (CSA), a type of population-based heuristic algo- preferentially to make up for inertial generation that renew-
rithm, have been presented by Hassanien [8]. Additionally, able energy source. The maximum frequency variation, sys-
the methods suggested in references [9, 10] are only tested tem inertia, and regulation constant are used to provide this
on the IEEE-30 bus system, which does not provide opti- security limitation. It is demonstrated that when this restric-
mal performance. M. Islam, et al., has been applied Salp tion is applied, the OPF preferentially chooses greater inertia
Swarm Algorithm (SSA) [11]. On the IEEE 30-bus standard generators in order to meet the frequency based security re-
system, Layth AL-Bahrani [12] suggested Improved Differ- quirement.While stability requirements in OPF have been in-
ential Evolution (IDE). Anbo Meng [13] suggests a unique vestigated previously, the frequency stability constraint has
crisscross search based grey wolf optimizer (CS-GWO) to not been mathematically modelled and implemented in the
address the OPF issue. The Salp Swarm Algorithm (ISSA), OPF issue.
developed by Salma Abd el-sattar [14], has been presented to OPF is one of the challenging issues that must be tack-
address the optimal power flow (OPF) problem by improv- led in the design and management of contemporary power
ing the search capabilities of the original SSA. The Heap systems. To achieve the highest level of security and relia-
optimisation algorithm (HOA) was utilised by Mohamed A. bility, it is expected that the energy system will need to run
M. Shaheen [15] to address the OPF issue. A unique hy- at maximum efficiency. In order to minimise the objective
brid multi-objective evolutionary algorithm (MOEA) based functions, it is necessary to determine the optimal control
on deconstruction and invasive weed optimisation (IWO) ap- variables for various power network components, including
proach was presented by Ravi Kumar Avvari [16] for the actual power generation, generator voltages, transformer set-
OPF problem. An improved ant lion optimisation (AALO) tings, reactive compensation components, etc. The majority
algorithm and the Lévy flight operator are combined in Mal- of power network issues related to overload and power qual-
lala Balasubbareddy’s [17] innovative ant lion optimisation ity can also be treated by a FACTS device. Since it has been
(ALO) algorithm. In order to prove the viability of the grey established that using FACTS devices improves the power
wolf optimisation approach (GWO), Inam Ullah Khan [18] performance of the electrical power network, finding the
used it to optimise modified IEEE-30 bus test systems. most efficient way to allocate FACTS devices to OPF issues.
Several research articles on OPF solutions with FACTS Numerous comparison studies on the effectiveness using al-
devices have been published during the last three decades. gorithms in solving engineering challenges have been looked
Sirote Khunkitti proposed Hybrid DA-PSO [19] to tackle the into and carried out in related papers. Studies of the litera-
OPF problem by adding FACTS. Douglas Rodrigues [20] ture suggest that there is growing interest in using new al-
Adaptive Improved Flower Pollination Algorithm (AIFP) for gorithms to address engineering issues, particularly in OPF
Global Optimization faster convergence rate. Pramod Ku- techniques. This research suggests employing the algorithm
mar Gouda [21] solve OPF problem using UPFC. In OPF, to address the OPF problem while also taking accounts the
Mohamed et al. [22] identify the proper ratings and locations existence of FACTS devices in an electrical network. In or-
for FACTS (TCSC/TCPS/SVC) devices. The OPF problem der to get over these restrictions, this study offers a novel,
in the IEEE-30 bus is analysed by Ankur Maheshwari [23], strong and well-proven new optimisation strategy termed as
accounting for solar PV, wind turbine, and (TES) tidal en- driving training-based optimization (DTBO). This strategy
ergy systems. According to Serhat Duman [24] research, has never been used or implemented before in an electric
test scenarios for RES are implemented using IEEE 30-bus power network using UPFC. The suggested method boosts
and 57-bus test systems to propose solutions for the security- the efficiency. The recommended algorithm’s search agents
constrained OPF. exhibit a variety of social and individual qualities to avoid

𝐶𝑜𝑟𝑟𝑒𝑠𝑝𝑜𝑛𝑑𝑖𝑛𝑔 𝑎𝑢𝑡ℎ𝑜𝑟 𝑚𝑎𝑖𝑙 𝑖𝑑: [email protected] Page 2 of 17

 Frequency Constrained OPF Incorporating UPFC Controller Using DTBO

2. Model: FACTS devices

2.1. Unified power flow controller (UPFC)
The static UPFC model and its mathematical represen-
tations are described below. The best adaptable FACTS tool
now existing for the control of transmission system. The
UPFC, which increases the power system’s transmission ca-
pacity, has the ability to regulate both active and reactive
load flow among its terminals. It can achieve fast-acting en-
ergy compensation in power network transient state, main-
taining voltage levels at the connecting point effectively. Ad-
ditionally, it could offer the node to which it is linked reac-
tive power compensation [25]. A common DC link connects
two voltage source converters, which make up the device. In
Fig.2, the UPFC circuit is depicted as being positioned be-
tween the 𝑒𝑡ℎ and 𝑓 𝑡ℎ buses. The following expressions can
be used to express the injected reactive and active power of
the 𝑒𝑡ℎ and 𝑓 𝑡ℎ bus.
Why using brackes before equation?

⎧ | || |
𝑃𝑒 = (𝐺𝑃 + 𝐺𝑆 )||𝑉𝑒 || − ||𝑉𝑒 || |𝐸𝑝 | |𝑌𝑝 | cos(𝜃𝑝 − 𝛿𝑒 + 𝛿𝑝 )+
2
Figure 1: The single line diagram of the IEEE-30 bus system ⎪ | || |
⎪ |𝑉𝑒 | |𝐸𝑠 | |𝑌𝑠 | cos(𝜃𝑠 + 𝛿𝑠 − 𝛿𝑒 ) − |𝑉𝑒 | ||𝑉𝑓 || |𝑌𝑠 | cos(𝜃𝑠 −
with thermal generator ⎨ | || || | | || || | (1)
⎪ ∑ | || || |
𝑁𝐵
⎪ 𝛿𝑒 + 𝛿𝑓 ) + |𝑉𝑒 | |𝑉𝑘 | |𝑌𝑒𝑘 | cos(𝜃𝑒𝑘 − 𝛿𝑒 + 𝛿𝑘 )
local optima entrapment and quicken convergence. ⎩ 𝑘=1

1.1. Motivation and incitement ⎧ 𝑄 = −(𝐵 + 𝐵 )|𝑉 |2 + |𝑉 | ||𝐸 || ||𝑌 || sin(𝜃 − 𝛿 + 𝛿 )−

⎪ 𝑒 𝑃 𝑆 | 𝑒| | 𝑒| | 𝑝| | 𝑝| 𝑝 𝑒 𝑝
The following research motivation factors are clarified in ⎪ ||𝑉𝑒 || ||𝐸𝑠 || ||𝑌𝑠 || sin(𝜃𝑠 + 𝛿𝑠 − 𝛿𝑒 ) + ||𝑉𝑒 || ||𝑉𝑓 || ||𝑌𝑠 || sin(𝜃𝑠 − 𝛿𝑒 + 𝛿𝑓 )
this article. ⎨ | | (2)
⎪ ∑ | || || |
𝑁𝐵
⎪ + 𝑉 𝑉 𝑌
| 𝑒 | | 𝑘 | | 𝑒𝑘 | sin(𝜃𝑒𝑘 − 𝛿 𝑒 + 𝛿 𝑘 )
a) The rising demand for power has made it more diffi- ⎩ 𝑘=1

cult for energy companies to efficiently and cheaply

balance supply and demand. ⎧ 𝑃 = +𝐺 ||𝑉 ||2 − ||𝑉 || |𝑉 | |𝑌 | cos(𝜃 − 𝛿 + 𝛿 ) − ||𝑉 || |𝐸 | |𝑌 |
⎪ 𝑓 𝑆| 𝑓 | | 𝑓 | | 𝑒| | 𝑠| 𝑠 𝑓 𝑒 | 𝑓 | | 𝑠| | 𝑠|
⎨ ∑ | || || |
𝑁𝐵 (3)
b) There are still several gaps in the research, as the lit- ⎪ cos(𝜃𝑠 + 𝛿𝑠 − 𝛿𝑒 ) + |𝑉𝑓 | |𝑉𝑘 | |𝑌𝑓 𝑘 | cos(𝜃𝑒𝑓 − 𝛿𝑒 + 𝛿𝑘 )
⎩ 𝑘=1 | | | |
erature study mentioned above shows. All of these
researcher did not concentrate about frequency devi-
⎧ | |2 | | | |
ation along with minimization of generation cost. In 𝑄𝑓 = −𝐵𝑆 |𝑉𝑓 | + ||𝑉𝑒 || |𝑉𝑓 | ||𝑌𝑠 || sin(𝜃𝑠 − 𝛿𝑓 + 𝛿𝑝 ) − |𝑉𝑓 | ||𝐸𝑠 || ||𝑌𝑠 ||
⎪ | | | | | |
⎨ ∑ | || || |
𝑁𝐵 (4)
this article concentrate about frequency deviation with ⎪ sin(𝜃𝑠 + 𝛿𝑠 − 𝛿𝑓 ) + |𝑉𝑓 | |𝑉𝑘 | |𝑌𝑒𝑓 | sin(𝜃𝑒𝑓 − 𝛿𝑓 + 𝛿𝑘 )
minimization of generation cost which is challenging ⎩ 𝑘=1 | | | |

work.
The following equations are reactive and active power
DTBO has been used in the current work to find the best over the transmission line existed between the 𝑒𝑡ℎ and 𝑓 𝑡ℎ
possible solution for an IEEE 30-bus OPF issue. Encasing bus having UPFC:
DTBO aims to avoid optimal local levels, accelerate conver- { | || |
𝑃𝑒𝑓 = (𝐺𝑃 + 𝐺𝑆 )||𝑉𝑒 || − ||𝑉𝑒 || |𝐸𝑝 | |𝑌𝑝 | cos(𝜃𝑝 − 𝛿𝑒 + 𝛿𝑝 )+
2
gence, increase outcome accuracy, and result in a notable in- | || | (5)
|𝑉𝑒 | |𝐸𝑠 | |𝑌𝑠 | cos(𝜃𝑠 + 𝛿𝑠 − 𝛿𝑖 ) − |𝑉𝑒 | ||𝑉𝑓 || |𝑌𝑠 | cos(𝜃𝑠 − 𝛿𝑒 + 𝛿𝑓 )
crease in OPF performance while reducing evaluation time. | || || | | || || |

An IEEE 30-bus network made up of conventional Thermal { | || |

𝑄𝑒𝑓 = −(𝐵𝑃 + 𝐵𝑆 )||𝑉𝑒 || + ||𝑉𝑒 || |𝐸𝑝 | |𝑌𝑝 | sin(𝜃𝑝 − 𝛿𝑒 + 𝛿𝑝 )−
2
Power units with UPFC is used as the test system. OPFs | || |
|𝑉𝑒 | |𝐸𝑠 | |𝑌𝑠 | sin(𝜃𝑠 + 𝛿𝑠 − 𝛿𝑖 ) + |𝑉𝑒 | ||𝑉𝑓 || |𝑌𝑠 | sin(𝜃𝑠 − 𝛿𝑒 + 𝛿𝑓 )
(6)
are designed to reduce voltage variations, emissions, power | || || | | || || |
loss, and total cost of generating. In the parts that follow, the
organisations of the three cases that are discussed are clearly {
| |2 | | | |
𝑃𝑓 𝑒 = +𝐺𝑆 |𝑉𝑓 | − |𝑉𝑓 | ||𝑉𝑒 || ||𝑌𝑠 || cos(𝜃𝑠 − 𝛿𝑓 + 𝛿𝑒 ) − |𝑉𝑓 | ||𝐸𝑠 || ||𝑌𝑠 ||
stated. | | | | | | (7)
cos(𝜃𝑠 + 𝛿𝑠 − 𝛿𝑒 )
Where this figure cited in the text?

⎧ | | | |
2
⎪ 𝑄𝑓 𝑒 = −𝐵𝑆 |𝑉𝑓 | + |𝑉𝑓 | ||𝑉𝑒 || ||𝑌𝑠 || sin(𝜃𝑠 − 𝛿𝑓 + 𝛿𝑝 )−
⎨ | | | | (8)
| || || |
⎪ |𝑉𝑓 | |𝐸𝑠 | |𝑌𝑠 | sin(𝜃𝑠 + 𝛿𝑠 − 𝛿𝑓 )
⎩ | |

𝐶𝑜𝑟𝑟𝑒𝑠𝑝𝑜𝑛𝑑𝑖𝑛𝑔 𝑎𝑢𝑡ℎ𝑜𝑟 𝑚𝑎𝑖𝑙 𝑖𝑑: [email protected] Page 3 of 17

 Frequency Constrained OPF Incorporating UPFC Controller Using DTBO

Figure 2: UPFC model placed between 𝑒𝑡ℎ bus and 𝑓 𝑡ℎ bus

It is possible to ignore the UPFC’s power losses. The defined what is g and h please
result is that there is no net active power exchange in the
where 𝑢 is the set of independent variables, also known
UPFC, which is stated as follows :
as control variables. The dependent variables, or state vari-
ables, is denoted by 𝑥. The control variables, 𝑢 and state
⎧ | |2 | | | | variables, 𝑥 of the OPF issue are specified as follows.
𝐺𝑝 |𝐸𝑝 | + 𝐺𝑠 ||𝐸𝑠 || − |𝐸𝑝 | ||𝑉𝑒 || |𝑌𝑝 | × cos(𝛿𝑝 − 𝛿𝑒 − 𝜃𝑝 ) + ||𝐸𝑠 || ||𝑉𝑒 || ||𝑌𝑠 ||
2
⎪
⎨ | | | | | | (9)
| |
⎪ × cos(𝛿𝑠 − 𝛿𝑒 − 𝜃𝑠 ) − ||𝐸𝑠 || |𝑉𝑓 | ||𝑌𝑠 || × cos(𝛿𝑠 − 𝛿𝑓 − 𝜃𝑠 ) = 0
⎩ | | 3.1. Control variables
These variables are modifiable to satisfy the load flow
where 𝑉𝑒 ,𝑉𝑓 indicate the magnitude of voltages at the calculations [26]. The OPF problem was formulated using
𝑒𝑡ℎ and 𝑓 𝑡ℎ buses, respectively; 𝑌𝑝 ,𝐵𝑝 and 𝐺𝑝 are the par- the following control variables:
allel component admittance, susceptance and conductance 𝑃𝐺 : real power at the PV buses except at the swing bus.
respectively; 𝑌𝑠 is the summed admittance of the series com- 𝑉𝐺 : magnitude of the voltage at the PV buses.
ponent of the UPFC and the line of transmission present be- T: tap changing transformer.
tween the e-f bus; 𝐺𝑠 ,𝐵𝑠 are the conductance and suscep- 𝑄𝐶 : shunt VAR compensator
tance of the UPFC series components respectively; 𝜃𝑒𝑘 is
[ ]
the admittance angle of connected transmission 𝑢𝑇 = 𝑃𝐺2 ⋯ 𝑃𝐺𝑁𝐺 , 𝑉𝐺1 ⋯ 𝑉𝐺𝑁𝐺 , 𝑄𝐶1 ⋯ 𝑄𝐶𝑁𝐶 , 𝑇1 ⋯ 𝑇𝑁𝑇 (13)
This ’ before 𝑒𝑡ℎ refers to what?
where NG, NT, and NC stand for the no. of generators, reg-
line between ’𝑒𝑡ℎ and 𝑘𝑡ℎ buses; 𝜃𝑠 is the admittance an- ulating transformers, and VAR compensators respectively.
gle between the admittance of the UPFC series component
and the admittance that contains the admittance of the line 3.2. State variables
’e-f’; 𝛿𝑝 and 𝛿𝑠 are the phase angle of the source voltage in These are the factors that distinguish each unique system
UPFC parallel and series elements respectively; 𝐸𝑝 ,𝐸𝑠 are state [26]. The following list of state variables is used to
the parallel and series converter’s sources of voltage in the formulate the OPF problem:
UPFC device respectively. 𝑃𝐺1 : at the slack bus real power.
𝑉𝐿 : voltage magnitude at the load buses.
3. Mathematical problem formulation 𝑄𝐺 : generation of reactive power from each generator
unit.
The OPF problem minimises a preset objective function, 𝑆1 : loading (or flow) of transmission lines. Thus, x can
such as the cost of active power generation and emission, be written as follows:
to determine the ideal control variable settings for a given
load. OPF is a nonlinear constrained optimisation problem [ ]
that takes into account the system’s operating restrictions. It 𝑋 𝑇 = 𝑃𝐺1 , 𝑉𝐿1 ⋯ 𝑉𝑁𝐿 , 𝑄𝐺1 ⋯ 𝑄𝐺𝑁𝐺 , 𝑆𝑙1 ⋯ 𝑆𝑙𝑛𝑙 (14)
can be stated as follows:
where the no. of transmission lines is represented by nl,
Minimize
and the no. of load buses by NL.
𝐽 (𝑥, 𝑢) (10)
3.3. Objective function
Subject to The usual OPF problem formulation incorporates multi-
ple valve steam turbines for flexible operating facilities, real
𝑔 (𝑥, 𝑢) = 0 (11) power generation, and generator terminal voltages in order to
And determine the best settings for control variables. The overall
generation cost for producing units with valve point discon-
ℎ (𝑥, 𝑢) ≤ 0 (12) tinuities is as follows:

𝐶𝑜𝑟𝑟𝑒𝑠𝑝𝑜𝑛𝑑𝑖𝑛𝑔 𝑎𝑢𝑡ℎ𝑜𝑟 𝑚𝑎𝑖𝑙 𝑖𝑑: [email protected] Page 4 of 17

 Frequency Constrained OPF Incorporating UPFC Controller Using DTBO

3.3.1. Total generation cost 3.6. Inequality constraints

(i) Generator constraints:
{
∑
𝑁𝐺
| ( ( ))|
𝐹𝐶 = 𝑎𝑖 + 𝑏𝑖 𝑃𝐺𝑖 + 𝑐𝑖 𝑃𝐺𝑖 2 + |𝑑𝑖 × sin 𝑒𝑖 × 𝑃𝐺𝑖
min − 𝑃
𝐺𝑖 || (15) ⎧ 𝑉𝐺𝑖
min ≤ 𝑉 max
| 𝐺𝑖 ≤ 𝑉𝐺𝑖
𝑖=1
⎪
⎪ min max
where 𝑎𝑖 , 𝑏𝑖 , 𝑐𝑖 , 𝑑𝑖 , 𝑒𝑖 : 𝑖𝑡ℎ generator cost coefficients; 𝑃𝐺𝑖 : ⎨𝑃𝐺𝑖 ≤ 𝑃𝐺𝑖 ≤ 𝑃𝐺𝑖 𝑖 ∈ 𝑁𝐺 (19)
𝑖𝑡ℎ generator real power generation; 𝑃𝐺𝑖 min : 𝑖𝑡ℎ generator low- ⎪
⎪ 𝑄min ≤ 𝑄 ≤ 𝑄max
est real power generation limit. ⎩ 𝐺𝑖 𝐺𝑖 𝐺𝑖

3.3.2. Emission (ii) Load bus constraints:

Use of fossil fuels to produce power emits harmful gasses
min max
into the environment. The emission from the thermal gener- 𝑉𝐿𝑖 ≤ 𝑉𝐿𝑖 ≤ 𝑉𝐿𝑖 𝑖 ∈ 𝑁𝐿 (20)
ators can be represented as:
(iii) Transmission line constraints:
⎧ ∑ −2 (
𝑁𝑇 𝐺 )
⎪𝑀𝑖𝑛2 = 10 𝜒𝑘 + 𝜂𝑘 𝑃𝑇 𝐺𝑘 + 𝜎𝑘 𝑃𝑇2𝐺𝑘 max
𝑆𝐿𝑖 ≤ 𝑆𝐿𝑖 𝑖 ∈ 𝑁𝑇 𝐿 (21)
⎨ ( 𝑘=1 ) (16)
⎪+ 𝜔 𝑒(𝜇𝑘 𝑃𝑇 𝐺𝑘 )
⎩ 𝑘 (iv) Transformer tap constraints:
Here 𝜒𝑘 , 𝜂𝑘 , 𝜎𝑘 , 𝜔𝑘 & 𝜇𝑘 are emission coefficients of 𝑘𝑡ℎ 𝑇𝑖min ≤ 𝑇𝑖 ≤ 𝑇𝑖max 𝑖 ∈ 𝑁𝑇 (22)
thermal generator.𝑃𝑇 𝐺𝑘 is the generated power by 𝑘𝑡ℎ ther-
mal generator. (v) Shunt compensator constraints:

3.3.3. Multi-objective 𝑄min max

𝐶𝑖 ≤ 𝑄𝐶𝑖 ≤ 𝑄𝐶𝑖 𝑖 ∈ 𝑁𝐶 (23)
It is very common that the above-mentioned objective
(minimization) functions are mutually conflicting with each where 𝑉𝐺𝑖 min , 𝑉 max : lowest and highest voltage limits respec-
𝐺𝑖
other. Often, it is required to obtain the best possible solu- tively of 𝑖𝑡ℎ generator bus; 𝑃𝐺𝑖 min , 𝑃 max : lowest and high-
𝐺𝑖
tion that optimizes those mutually-opposing objectives at a est limitations on active power generation respectively of 𝑖𝑡ℎ
time without violating various constraints. These types of bus; 𝑄min , 𝑄max : lowest and highest um reactive power gen-
optimization issues are referred to as multi-objective opti- 𝐺𝑖 𝐺𝑖
eration limits respectively of 𝑖𝑡ℎ bus; 𝑉𝐿𝑖 min , 𝑉 max : lowest and
mization problems. 𝐿𝑖
max :apparent power
highest voltage of 𝑖𝑡ℎ load bus; 𝑆𝐿𝑖 , 𝑆𝐿𝑖
3.3.3.1. Simultaneously Cost and Emission minimiza- flow and the maximum allowable apparent power flow of 𝑖𝑡ℎ
tion: Mathematically the function can be written as: branch;𝑇𝑖min , 𝑇𝑖max : limits of lowest and highest tap setting
𝑂𝐹𝑐𝑜𝑚𝑏1 = 𝐹 𝐶 + 𝛼𝐸𝐶 𝐹 2 (17) respectively of 𝑖𝑡ℎ tap changing transformer; 𝑄min 𝐶𝑖
, 𝑄max
𝐶𝑖
:
lowest and highest reactive power injection limits respec-
𝛼𝐸𝐶 represents weighting factor corresponding to emission. tively of 𝑖𝑡ℎ shunt compensator; 𝑁𝐺: no. of generator buses;
𝑁𝐿: no. of load buses; 𝑁𝑇 𝐿: no. of transmission line;
3.4. Constraints 𝑁𝑇 : no. of tap changer; 𝑁𝐶: no. of shunt compensator.
The OPF is bound by the following restrictions.
(vi) Frequency based security constraint:
where these restrictions? if you mean next one so I A power system experiences an unexpected disturbance
think we need to renumber them to subsubsection when the frequency deviates from its nominal value. By us-
ing load damping, inertia, and other damping methods, the
3.5. Equality constraints frequency fluctuation is quickly reduced. To prevent more
The load flow equations are express by these constraints frequency deviation, the governor adjusts the generator 𝑜∕𝑝.
as follows: It is possible to change the generator’s 𝑜∕𝑝 by modifying the
⎧∑ ( ) 𝑁𝐵
∑ 𝑁𝐵 ∑ | || || | ( ) prime mover 𝑖∕𝑝. The greatest frequency deviation and a
𝑁𝐵
⎪ 𝑃𝐺𝑖 − 𝑃𝐿𝑖 = |𝑉𝑖 | ||𝑉𝑗 || ||𝑌𝑖𝑗 || Cos 𝜃𝑖𝑗 − 𝛿𝑖𝑗 portion of the frequency recovery time are caused by these
⎪ 𝑖=1 𝑖=1 𝑗=1
acts. If the system has enough reserve, this procedure is
⎨ (18)
⎪ 𝑁𝐵
∑ ∑ 𝑁𝐵
𝑁𝐵 ∑ | || || | ( ) part of the LFC, which raises the frequency gradually until
⎪ 𝑄 − 𝑄𝐿𝑖 = − |𝑉𝑖 | ||𝑉𝑗 || ||𝑌𝑖𝑗 || Sin 𝜃𝑖𝑗 − 𝛿𝑖𝑗 it reaches its nominal value. The frequency deviation has an
⎩ 𝑖=1 𝐺𝑖 𝑖=1 𝑗=1
inverse relationship with load damping (D) and system iner-
𝑃𝐿𝑖 : active power demand of 𝑖𝑡ℎ bus; 𝑄𝐺𝑖 , 𝑄𝐿𝑖 : genera- tia (H). Thus, for the system to keep the frequency within a
tion and consumption of reactive power respectively of 𝑖𝑡ℎ safe range, a minimal amount of inertia is required. Because
bus;𝑖𝑡ℎ : real and reactive power injected respectively at 𝑖𝑡ℎ of this, the OPF problem should take a frequency based secu-
bus; 𝑃𝑖𝑠 , 𝑄𝑖𝑠 : admittance of the linked transmission line be- rity limitation into account. [27] contained a proposed gen-
tween 𝑖𝑡ℎ bus and 𝑗 𝑡ℎ bus; 𝜃𝑖𝑗 : angle of admittance of the eralised model of LFC that may explain the role played by
linked transmission line between 𝑖𝑡ℎ bus and 𝑗 𝑡ℎ bus; 𝑁𝐵 : each governor in the system frequency control. Based on the
no. of buses. low-order LFC model in [28], whose correctness has been

𝐶𝑜𝑟𝑟𝑒𝑠𝑝𝑜𝑛𝑑𝑖𝑛𝑔 𝑎𝑢𝑡ℎ𝑜𝑟 𝑚𝑎𝑖𝑙 𝑖𝑑: [email protected] Page 5 of 17

 Frequency Constrained OPF Incorporating UPFC Controller Using DTBO

( )
12𝐻 + 𝑇𝑅 𝐷 + 𝐹𝑅
𝜍= √ ( ) (29)
2 2𝐻𝑇𝑅 𝐷 + 𝑅𝑅

For Fig 25 the inverse Laplace transform is made possible

by the following expansion:
Δ𝑃𝐿 1 Δ𝑃𝐿
Δ𝑓 (𝑠) = 2𝐻𝑇𝑅 𝑠 𝑠2 +2𝜍𝜔𝑛 𝑠+𝜔2𝑛
+ 2𝐻 (𝑠2 +2𝜍𝜔𝑛 𝑠+𝜔2𝑛 )
(30)

The frequency deviation equation in the time domain can be

derived as follows from (22):
Figure 3: LFC model of a multi-machine system ( ( √ ))
⎧ Δ𝑃𝐿 1 −𝜍𝜔𝑛 𝑡 Cos 𝜔 2𝑡 − 𝜙
⎪ Δ𝑓 (𝑡) = 1 − √ 𝑒 𝑛 1 − 𝜍
2𝐻𝑇𝑅 𝜔2𝑛 2
assessed, this LFC model was built by measuring the fre- ⎨ ( 1−𝜍√ ) (31)
Δ𝑃𝐿
⎪+ √ 𝑒 −𝜍𝜔 𝑛 𝑡 Sin 𝜔𝑛 1 − 𝜍 𝑡 2
quency deviation’s sensitivity to both governor parameters. ⎩ 2𝐻𝜔𝑛 1−𝜍 2
The multi-machine system’s LFC model is shown in Fig. 3.
where
In Fig.3 the following notations are employed:
( )
H: corresponding constant of inertia; 𝜍
D: constant of load damping; 𝜙 = tan−1 √ (32)
𝐾𝑚 : LFC controller of 𝑚𝑡ℎ machine; 1 − 𝜍2
𝑅𝑚 : Corresponding regulation constant of 𝑚𝑡ℎ machine.
The derivative of the frequency deviation equals zero when
𝐹𝑚 : fractional power of 𝑚𝑡ℎ turbine;
it reaches its highest value:
𝑇𝑚 : time constant of governor of machine m;
Δ𝑓 : deviation of frequency; 𝑑Δ𝑓 (𝑡)
Δ𝑃𝐿 : disturbances. =0 (33)
𝑑𝑡
The inertia constant H in this specific model is the ra-
tios between the kinetic energy and the rating of synchronous Therefore, the following can be used to derive the greatest
machine. It is assumed that the values of all governor time frequency deviation and the associated time instant:
constants are equal. Because of the maximum frequency ( √ )
1 −1 𝜔𝑛 1 − 𝜍 2
deviation’s low sensitivity to the governor time constant 𝑡max = √ tan (34)
(TR), this assumption is fair and has minimal impact on the 𝜔 1 − 𝜍2 𝜍𝜔𝑛 − 1 ∕𝑇𝑅
𝑛
model’s correctness [27]. The frequency deviation equation
can be created as follows using the LFC model for the multi √ ( )
system: Δ𝑃𝐿 ⎛⎜ 𝑇𝑅 𝑅𝑅 − 𝐹𝑅 ⎞
⎟ (35)
Δ𝑓max = 1 + 𝑒𝜍𝜔𝑛 𝑡max
𝑅𝑅 + 𝐷 ⎜ 2𝐻 ⎟
Δ𝑃𝐿 ⎝ ⎠
𝑠
Δ𝑓 (𝑠) = (24)
∑𝑚 𝐾𝑖 (1+𝐹𝑖 𝑇𝑅 𝑠) The frequency based security restriction is created based on
𝐷 + 2𝐻𝑠 + 𝑖=1 𝑅𝑖 (1+𝑇𝑅 𝑠) the need to keep the frequency deviation within certain safe
bounds after the maximum frequency deviation equation has
Rewriting 24 as [27] is possible if all TR values for the sys- been established:
tem governors are assumed to be the same.
Δ𝑃𝐿 1 + 𝑇𝑅 𝑠 −Δ𝑓𝑠 ≤ Δ𝑓max ≤ Δ𝑓𝑠 (36)
Δ𝑓 (𝑠) = (25)
2𝐻𝑇𝑅 𝑠 𝑠2 + 2𝜍𝜔𝑛 𝑠 + 𝜔2𝑛 The maximum frequency deviation will fluctuate accord-
where ing to the generators’ variations in inertia. The inertia of
generator 𝑖 is included in the overall inertia of the system if
∑
𝑚
𝐾𝑖 𝐹𝑖 it is synchronised, which indicates that generator 𝑖 dispatches
𝐹𝑅 = (26)
𝑅𝑖 in the system. Otherwise, generator 𝑖’s inertia is not taken
𝑖=1
into account. The inertia then, in eq 26 and eq 27, the inertia
and analogous regulation constants are substituted. The way
∑
𝑚
𝐾𝑖 these two components change will have an impact on Δ𝑓max .
𝑅𝑅 = (27)
𝑖=1
𝑅𝑖 As demonstrated in eq 35, the frequency deviation increases
when inertia is reduced and decreases in other cases. There-
√ fore, in the OPF problem, eq 36 is employed as the security
1 ( ) constraint. The suggested solution takes into account a load
𝜔𝑛 = 𝐷 + 𝑅𝑅 (28)
2𝐻𝑇𝑅 increase as the frequency disturbance, but it can also handle

𝐶𝑜𝑟𝑟𝑒𝑠𝑝𝑜𝑛𝑑𝑖𝑛𝑔 𝑎𝑢𝑡ℎ𝑜𝑟 𝑚𝑎𝑖𝑙 𝑖𝑑: [email protected] Page 6 of 17

 Frequency Constrained OPF Incorporating UPFC Controller Using DTBO

generation tripping. The system’s generator (let’s say gener- where 𝑧max min
𝑝𝑞 and 𝑧𝑝𝑞 are the upper and lower bounds of
ator 𝑖) will stop contributing to load frequency control once the 𝑞 𝑡ℎ variable; 𝑟 is the unbiased random number which lies
it trips (𝐻𝑖 , 𝑅𝑖 , 𝐹𝑖 , 𝑇𝑖 , 𝐾𝑖 ). In this instance, the frequency between 0 and 1.
based security constraint uses this value of frequency, which The objective function’s value is calculated for each
is obtained by changing the maximum frequency deviation unique candidate solution and is shown as follows:
in eq 35. The suggested approach will still be used, and more
units will be involved in the dispatch process to ensure fre- ( )
quency based security. ⎡⎧ 𝐹1 ⎤ ⎡⎧ 𝐹 𝑍1 ⎤
⎢⎪ ⎥ ⎢⎪ ⎥
The suggested approach can be used in both systems that ⎢⎪ . ⎥ ⎢⎪ ⎥
incorporate renewable energy resources (RERs) and systems ⎢⎪ . ⎥ ⎢⎪ ⎥
that simply have conventional generation. The limited abil- ⎢⎪ ⎥ ⎢⎪ ( ) ⎥
𝐹 = ⎢⎨ 𝐹𝑝 ⎥ = ⎢⎨ 𝐹 𝑍𝑝 ⎥ (39)
ity of RERs to assist frequency control systems from their ⎢⎪ . ⎥ ⎢⎪ ⎥
near-zero inertia. However, when its available output devi- ⎢⎪ ⎥ ⎢⎪ ⎥
ates from zero, RERs can simulate a conventional generator ⎢⎪ . ⎥ ⎢⎪ ⎥
⎢⎪ ⎥ ⎢⎪ ( )⎥
with its low synthetic inertia by implementing specific com- ⎣⎩ 𝐹𝑁 ⎦𝑁×1 ⎣⎩ 𝐹 𝑍𝑁 ⎦𝑁×1
plicated control algorithms [29].The suggested approach is
therefore still in use. The computed values of the objective function serve as the
deciding factor in assessing the qualities of the solutions that
are being considered. The candidate solution that produces
4. Optimization Algorithm the optimal value of the goal function is chosen as the best
4.1. DTBO member. The top member is modified as the iteration pro-
Dehghani 𝑒𝑡 𝑎𝑙. introduced DTBO as in [30]. DTBO ceeds. The DTBO candidate solution revision process con-
is a population-based meta-heuristic technique. The DTBO sists of the following three steps.:
program emulates the manner in which a driving instruc-
Step 1: Driving teacher training (Exploration): The major-
tor instructs trainees in a driving school. The mathemati-
ity of the DTBO community is categorised as trainee
cal framework of DTBO contains three phases: (1)instruc-
drivers, with just a small number of attractive individ-
tion provided by the driving instructor, (2) student patterning
uals chosen to be driving instructors. The process of
based on instructor expertise, and (3) practice. The ability
expertly selecting teachers and developing their tal-
of novice drivers to learn and master the skill of driving de-
ents as educators allows them to explore the globe for
pends on their level of intelligence. A seasoned driver can
the ideal place to solve a specific problem. Every it-
learn from a variety of instructors in driving school. Driv-
eration, instructors are chosen from among 𝐿 DTBO
ing skills are developed by new drivers through practicing on
members based on a comparison of the values of the
their own and by according to their instructor’s instructions.
objective function. These members are represented as
These learner-teacher exchanges and the self-practice of en-
the driving matrix 𝐷𝐼 in the following manner:
hancing driving abilities form the basis of the mathematical
modelling of DTBO.The DTBO population matrix is shown
below, with each row member representing a potential solu- ⎡⎧ 𝐷𝐼11 . . 𝐷𝐼1𝑞 . 𝐷𝐼1𝑚 ⎤
⎡⎧ 𝐷𝐼1 ⎤ ⎢⎪ ⎥
tion to the given problem: ⎢⎪ . ⎥ ⎢⎪ . . . . . ⎥
⎢⎪ ⎥ ⎢⎪ . . . . ⎥
⎢⎪ . ⎥ ⎢⎪ ⎥
𝐷𝐼 = ⎢⎨ ⎥ = ⎢⎨ 𝐷𝐼𝑝1 . . 𝐷𝐼𝑝𝑞 . 𝐷𝐼𝑝𝑚 ⎥ (40)
⎢⎪ 𝐷𝐼𝑝 ⎥ ⎢⎪ ⎥
⎡⎧ 𝑍1 ⎤ ⎡⎧ 𝑧11 . . 𝑧1𝑞 . 𝑧1𝑚 ⎤ . . . . .
⎢⎪ ⎥ ⎢⎪ ⎥
⎢⎪ ⎥ ⎢⎪ ⎥ ⎢⎪ ⎥ ⎢⎪ . . . . . ⎥
⎢⎪ . ⎥ ⎢⎪ . . . . . . ⎥ ⎣⎩ 𝐷𝐼𝐿 ⎦𝐿×𝑚 ⎢⎪ ⎥
⎢⎪ ⎣⎩ 𝐷𝐼𝐿1. . . 𝐷𝐼𝐿𝑞 . 𝐷𝐼𝐿𝑚 ⎦𝐿×𝑚
. ⎥ ⎢⎪ . . . . . . ⎥
⎢⎪ ⎥ ⎢⎪ ⎥
𝑍 = ⎢⎨ 𝑍𝑝 ⎥ = ⎢⎨ 𝑧𝑝1 . . 𝑧𝑝𝑞 . 𝑧𝑝𝑚 ⎥ (37)
⎢⎪ 𝐷𝐼𝑝 is 𝑝𝑡ℎ driving instructor. 𝐷𝐼𝑝𝑞 is 𝑞 𝑡ℎ variable of
. ⎥ ⎢⎪ . . . . . . ⎥
⎢⎪ ⎥ ⎢⎪ ⎥ 𝑝𝑡ℎ instructor.
⎢⎪ . ⎥ ⎢⎪ . . . . . . ⎥
⎢⎪ ⎥ ⎢⎪ ⎥ ⌊ ( )⌋
⎣⎩ 𝑍𝑁 ⎦𝑁×𝑚 ⎣⎩ 𝑧𝑁1. . . 𝑧𝑁𝑞 . 𝑧𝑁𝑚 ⎦𝑁×𝑚 1−𝑠
𝐿 = 0.1 × 𝑁 × (41)
𝑆
The DTBO population is indicated by Z; 𝑝𝑡ℎ member of Z is
𝑍𝑝 ; 𝑧𝑝𝑞 is the 𝑞 𝑡ℎ variable of the 𝑝𝑡ℎ solution;The population S is the maximum iteration, while s represents the cur-
size is 𝑁; No of problem variables is indicated by 𝑚. rent iteration. The adjusted position of the DTBO pop-
The starting positions of DTBO members are initialized ulation member is obtained as follows in this step:
at random at the start of DTBO implementation in the fol-
lowing ways: { ( )
𝑧𝑝𝑞 + 𝑟. 𝐷𝐼𝑘𝑝𝑞 − 𝐼.𝑧𝑝𝑞 , 𝐹𝐷𝐼𝑘𝑝 < 𝐹𝑝
𝑧𝑠𝑡1
𝑝𝑞 = ( ) (42)
( ) 𝑧𝑝𝑞 + 𝑟. 𝑧𝑝𝑞 − 𝐷𝐼𝑘𝑝𝑞 , 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒
𝑧𝑝𝑞 = 𝑧min + 𝑟 ∗ 𝑧 max − 𝑧min 𝑓 𝑜𝑟 𝑝 = 1 𝑡𝑜 𝑁 𝑞 = 1 𝑡𝑜 𝑚
𝑝𝑞 𝑝𝑞 𝑝𝑞 In the set {1, 2}, 𝐼 represents a random no, and 𝑟 rep-
(38) resents a random value between 0 and 1. A random

𝐶𝑜𝑟𝑟𝑒𝑠𝑝𝑜𝑛𝑑𝑖𝑛𝑔 𝑎𝑢𝑡ℎ𝑜𝑟 𝑚𝑎𝑖𝑙 𝑖𝑑: [email protected] Page 7 of 17

 Frequency Constrained OPF Incorporating UPFC Controller Using DTBO

selection of 𝑘 is made from the collection 1,2,...,L in

𝐷𝐼𝑘𝑝𝑞 𝑖.𝑒. 𝑘𝑡ℎ driving instructor whose objective func- ( )
𝑠
tion value is 𝐹𝐷𝐼𝑘𝑝 , 𝑝 denotes 𝑝𝑡ℎ trainee member of 𝑧𝑠𝑡3
𝑝,𝑞 = 𝑧 𝑝𝑞 + (1 − 2𝑟) .𝑅. 1 − .𝑧𝑝𝑞 (47)
𝑆
the population which is under the training of 𝑘𝑡ℎ in-
structor. {
When new position provides fitter solution than earlier 𝑍𝑝𝑠𝑡3 , 𝐹𝑝𝑠𝑡3 < 𝐹𝑝
𝑍𝑝 = (48)
position then the position is updated by (43). 𝑍𝑝 , 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒

{ 𝑍𝑝𝑠𝑡3 is modified 𝑝𝑡ℎ possible solution at the 3𝑟𝑑 DTBO

𝑍𝑝𝑠𝑡1 , 𝐹𝑝𝑠𝑡1< 𝐹𝑝 phase; 𝑧𝑠𝑡3 𝑡ℎ
𝑝,𝑞 is its 𝑞 variable; the value of the related
𝑍𝑝 = (43)
𝑍𝑝 , 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒 objective function is 𝐹𝑝𝑠𝑡3 ; 𝑟 is arbitrary quantity, rang-
ing from 0 to 1.; 𝑅 is 0.05, 𝑠 is present iteration and 𝑆
The revised 𝑝𝑡ℎ candidate solution at the 1𝑠𝑡 DTBO is the maximum iteration.
step is 𝑍𝑝𝑠𝑡1 ; 𝑧𝑠𝑡1 𝑡ℎ
𝑝𝑞 is its 𝑞 problem variable, The value Steps one through three update the DTBO population,
of its objective function is 𝐹𝑝𝑠𝑡1 . completing one DTBO iteration. Then, with a freshly
updated population, the subsequent iteration begins
and this procedure is ongoing [through (40) to (48)]
Step 2: The student driver’s teacher skills are patterned till the end of the last iteration. The best potential so-
(exploration): In the 2𝑛𝑑 step, the trainee driver mim- lution is noted as the problem’s solution at the conclu-
ics the instructor’s techniques and actions to enhance sion of the last iteration. Flowchart of DTBO is shown
the DTBO solutions. Members of the DTBO reach in Fig. 4
a new area of the search space through this proce-
dure. It strengthens DTBO’s exploration power. The you can add section for implementation or steps of de-
DTBO members and instructors combine linearly to veloped model or update DTBO section to contain in-
form a modified position, which is mathematically formation related to problems and objective functions
represented by (44). The previous position is replaced
by (45) if the goal function’s value is higher at the new
location than it was at the old one. 5. Results of simulation and comparisons for
different cases
The current section shows the results of the OPF simula-
𝑧𝑠𝑡2
𝑝𝑞 = 𝜉.𝑧𝑝𝑞 + (1 − 𝜉) .𝐷𝐼𝑘𝑝𝑞 (44) tions in which there are three test system to characterise the
overall investigation. In each test system three cases are con-
sidered. Table 1 shows the three test systems with various
{ cases. The IEEE-30 bus system is taken into consideration
𝑍𝑝𝑠𝑡2 𝐹𝑝𝑠𝑡2 < 𝐹𝑝
𝑍𝑝 = (45) here, which includes 41 transmission lines, 6 generators, 4
𝑍𝑝 , 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒 tap changers and 9 compensating devices and total demand
for 283.4 MW of active load and 126.2 MVAR of reactive
The 𝑍𝑝𝑠𝑡2 is the updated 𝑝𝑡ℎ candidate solution on the load as specified in Table 4. The frequency deviation and
DTBO second stage, 𝑧𝑠𝑡2 𝑡ℎ
𝑝𝑞 is its 𝑞 variable, The re-
generating cost are seen when the load fluctuates between
lated objective function value is 𝐹𝑝𝑠𝑡2 . The patterning 50% and 100%. The main objective is to minimize the gen-
index 𝜉 is given by: eration cost, emission, and frequency deviation at various
loads. Frequency deviation is inversely proportional to the
inertia constant as previously explained. The cost and emis-
( )
𝑠 sion coefficients for thermal generators are displayed in Ta-
𝜉 = 0.01 + 0.9 1 − (46)
𝑆 ble 2. Table 5 shows that the generation cost, emission, ac-
tive power loss, voltage profile and frequency deviation at
Step 3: Personal practice (Exploitation): In this phase, the various load(50% to 100%) for case 1.
driving skills of the inexperienced drivers are en- At 50% load for case 1, the output power of Thermal
hanced through individual practice. It’s similar to tak- plants PTG8, PTG11, and PTG13 becomes zero. Thus, out
ing advantage of DTBO’s local search feature. Ev- of 6 power plants, only 3 (PTG1, PTG2, and PTG5) are op-
ery student searches for a better position that comple- erating at this load. Therefore sum of the inertia of all gener-
ments their current one. New positions are created ators will decrease, and the inertia constant at various gen-
close to the current position by (47). Using (48), the eration powers are shows in Fig. 5. Since frequency de-
old position is swapped out for the new one, and the viation and inertia constant are inversely co-related, under
goal function value is upgraded in the manner shown this load, generation cost is minimal but frequency devia-
below: tion is largest because of the decrease in inertia constant.

𝐶𝑜𝑟𝑟𝑒𝑠𝑝𝑜𝑛𝑑𝑖𝑛𝑔 𝑎𝑢𝑡ℎ𝑜𝑟 𝑚𝑎𝑖𝑙 𝑖𝑑: [email protected] Page 8 of 17

 Frequency Constrained OPF Incorporating UPFC Controller Using DTBO

Figure 4: Flowchart of DTBO

Also, for 60% of the load, the same things occur, only three increasing load but increasing generation cost.
generators—PTG1, PTG2, and PTG5—are operating under Since frequency deviation is a crucial component of the
this load; the other three—PTG8, PTG11, and PTG13—do generating system, the primary aim is to minimise its im-
not operate at all. Consequently, the frequency deviation pact on all loads while simultaneously minimising the gen-
rises with these loads as well. However, it has been seen eration cost and emission. In case1 frequency based se-
that all generators are running at higher loads 𝑖.𝑒 80%, 90%, curity constraints are not taken into account. The crucial
and 100%. Therefore, frequency deviation decreases with parameter, the frequency based security constraint, is now

𝐶𝑜𝑟𝑟𝑒𝑠𝑝𝑜𝑛𝑑𝑖𝑛𝑔 𝑎𝑢𝑡ℎ𝑜𝑟 𝑚𝑎𝑖𝑙 𝑖𝑑: [email protected] Page 9 of 17

 Frequency Constrained OPF Incorporating UPFC Controller Using DTBO

Table 1 Table 5
Various case-studies investigated in this article Solution of the OPF for variable load without the frequency
Case
Single Multi
Considered objectives Constraints Test system security restriction for the IEEE 30 bus test system (Case 1)
objective objective
√ IEEE 30 Bus without
1
Total Cost minimization with valve
Equality and in-equality the frequency based
Control
point effects for thermal power plant Min. Max. 50% 60% 70% 80% 90% 100%
√ security constraint. Parameters
2 Emission and carbon tax minimization Equality and in-equality
√ Simultaneous minimization of Cost
PTG1(MW) 0 20082.34 150.78 169.81 169.81 169.63 169.8
3 Equality and in-equality
with Emission and carbon tax PTG2(MW) 0 80 43.01 11.98 16.88 36.57 52.64 72.65
IEEE 30 Bus with fre-
4
√ Total Cost minimization with valve
Equality and in-equality quency based security
PTG5(MW) 0 50 19.6 11.91 13.37 19.96 18.91 18.07
point effects for thermal power plant PTG8(MW) 0 35 0 0 0 1.15 12.22 10.54
√ constraint.
5 Emission and carbon tax minimization Equality and in-equality PTG11(MW) 0 30 0 0 2.08 3.73 4.35 11.73
√ Simultaneous minimization of Cost
6 Equality and in-equality PTG13(MW) 0 40 0 0 2.12 2.8 6.07 9.83
with Emission and carbon tax
√ IEEE 30 Bus with fre- 𝑉1 (pu) 0.95 1.11.0499 1.0717 1.0864 1.0968 1.0553 1.098
Total Cost minimization with valve
7 Equality and in-equality quency based security
√
point effects for thermal power plant
constraint and UPFC 𝑉2 (pu) 0.95 1.11.0548 1.0529 1.0708 1.0869 1.0475 1.0819
8 Emission and carbon tax minimization Equality and in-equality 𝑉5 (pu) 0.95 1.11.078 1.0296 1.0428 1.0605 1.023 1.0517
√ Simultaneous minimization of Cost
9
with Emission and carbon tax
Equality and in-equality 𝑉8 (pu) 0.95 1.11.0254 1.0353 1.0457 1.0683 1.0289 1.0582
𝑉11 (pu) 0.95 1.11.0043 1.0671 1.0872 1.0979 1.0823 1.099
𝑉13 (pu) 0.95 1.11.0355 1.0857 1.0931 1.0998 1.0446 1.0982
𝑇11 (pu) 0.9 1.10.9238 1.0008 0.9865 1.0056 1.0749 1.0585
Table 2 𝑇12 (pu) 0.9 1.11.0883 0.9444 0.9901 1.0199 0.939 0.9
Cost and emmission coefficients of thermal generators for IEEE 𝑇15 (pu) 0.9 1.11.0091 0.9926 1.0068 1.0326 0.9775 0.9989
𝑇36 (pu) 0.9 1.10.9719 0.9709 0.9732 0.9993 0.968 0.964
30-bus test system 𝑄𝐶10 (MVAR) 0 0.05
0.0343 0.0482 0.047 0.0495 0.0166 0.0496
Generator Bus 𝑎 𝑏 𝑐 𝑑 𝑒 𝛼 𝛽 𝛾 𝜔 𝜇 𝑄𝐶12 (MVAR) 0 0.05
0.0289 0.0481 0.0458 0.048 0.0493 0.0494
TG1 1 0 2 0.00375 18 0.037 4.091 -5.554 6.49 0.0002 2.857 𝑄𝐶15 (MVAR) 0 0.05
0.0466 0.0217 0.0311 0.0383 0.042 0.0483
TG2 2 0 1.75 0.0175 16 0.038 2.543 -6.047 5.638 0.0005 3.333 𝑄𝐶17 (MVAR) 0 0.05
0.0369 0.0391 0.0453 0.0484 0.0421 0.0492
TG5 5 0 1 0.0625 14 0.04 4.258 -5.0494 4.586 0.000001 8 𝑄𝐶20 (MVAR) 0 0.05
0.0377 0.0231 0.0297 0.03 0.04 0.0477
TG8 8 0 3.25 0.00834 12 0.045 5.326 -3.55 3.38 0.002 2
𝑄𝐶21 (MVAR) 0 0.05
0.0327 0.0485 0.0485 0.0498 0.0495 0.0489
TG11 11 0 3 0.025 13 0.042 4.258 -5.094 4.586 0.000001 8
TG13 13 0 3 0.025 13.5 0.041 6.131 -5.555 5.151 0.00001 6.667 𝑄𝐶23 (MVAR) 0 0.05
0.0455 0.0086 0.0103 0.0204 0.0257 0.0326
𝑄𝐶24 (MVAR) 0 0.05
0.0187 0.0469 0.0493 0.0477 0.0484 0.0499
𝑄𝐶29 (MVAR) 0 0.05
0.0158 0.0179 0.0182 0.0219 0.0182 0.0285
Total generation 144.95 175.28 205.02 234.02 263.82 292.62
Table 3 Total generation cost
($∕ℎ) 368.9182 458.4223 541.5061 633.8415 738.8719 843.5678
Emission((t/h) 0.2434 0.3206 0.3448 0.3352 0.3288 0.3268
Power, Voltage limits and Dynamic data of thermal generators 𝑃𝑙𝑜𝑠𝑠 (MW) 3.25 5.24 6.63 7.29 8.76 9.22
for a IEEE 30-bus test system VD(pu) 0.8405 1.3951 1.5382 1.5685 0.6605 1.6701
Frequency deviation with load
Generator Bus 𝑃𝑚𝑖𝑛 𝑃𝑚𝑎𝑥 𝑄𝑚𝑖𝑛 𝑄𝑚𝑎𝑥 𝑉𝑚𝑖𝑛 𝑉𝑔 H % max load OS SS(Sec)
TG1 1 50 200 -20 150 0.95 1.05 5.52 100 0.009890 17.02
Frequency
Deviation

TG2 2 20 80 -20 60 0.95 1.05 2.28 90 0.009801 11.21

80 0.009886 8.34
TG5 5 15 50 -15 62.5 0.95 1.05 1.75
70 0.009838 8.71
TG8 8 10 35 -15 48.7 0.95 1.05 1.15 60 0.009896 7.68
TG11 11 10 30 -10 40 0.95 1.05 0.95 50 0.01032 6.37
TG13 13 12 40 -15 44.7 0.95 1.05 1.35

Table 4 (OS=0.01004 and SS=6.3Sec). In case 4,when frequency

An overview of IEEE 30- bus System under study based security constraints are taken into account,the over-
Items Quantity Details
all generating cost at the same load is 369.5837 ($/h) which
Buses 30 [31] is higher as compare to case 1 but frequency deviation
Branches 41 [31]
Thermal generators 6 Buses:1 (swing), 2, 5, 8, 11 and 13 (OS=0.0098 and SS=5.01Sec) is reduced. However in case
Tap changing transformer 4
Branches:(6-9), (6-10), (4-12) and (25- 7, when introduced UPFC, generation cost is reduced to
27)
Active power for Generators (5Nos.), 367.3087 ($/h) which is less than case 1 and case 4 while
Control variables 24
voltages of all generator buses (6Nos.), also reduced the frequency deviation (OS=0.009709 and
transformer tap changing (4Nos.) and
compensation devices (9Nos.) SS=4.98) and power loss. The generating cost, emission,
Load demand 283.4MW, 126.2MVAR active power loss, voltage profile, and frequency deviation
Range of load bus voltage 24 [0.95-1.05]𝑝𝑢
Buses:10, 12, 15, 17, 20, 21, 23, 24 and for case 7 (incorporating UPFC and frequency based secu-
Compensation devices 9
29 rity constraint) is displayed in Table 7.
When the objective is to minimize the emission at
50% load for case 2, emission is minimum 𝑖.𝑒 0.2269
taken into consideration.Table 6 shows that the generat-
(t/h) but frequency deviation is maximum (OS=0.01083
ing cost,emission,active power loss,voltage profile, and fre-
and SS=6.42Sec) which is displayed in Table 8. In case
quency deviation under different loads (50% to 100%) for
5,when frequency based security constraints are taken into
case 4. Frequency based security constraint means the range
account,emission at the same load is 0.213 (t/h) which is less
of minimum and maximum power of generators will take in
as compare to case 2 but frequency deviation (OS=0.010046
such a way that all generators will run at any load. If all gen-
and SS=5.809) is reduced which is displayed in Table 9.
erators are run simultaneously,the frequency deviation will
However in case 8, when introduced UPFC, emission is re-
be reduced because of the increased sum of inertia constants
duced to 0.2127 (t/h) which is less than case 2 and case 5
but the generating costs will be increases due to run all gen-
while also reduced the frequency deviation (OS=0.0100449
erators.
and SS=5.79) and power loss. The generating cost, emis-
When the objective is to minimize the generation cost
sion, active power loss, voltage profile, and frequency devi-
at 50% load for case 1, total generation cost is minimum
ation for case 8 (incorporating UPFC and frequency based
𝑖.𝑒 368.9182 ($/h) but frequency deviation is maximum
security constraint) is displayed in Table 10.

𝐶𝑜𝑟𝑟𝑒𝑠𝑝𝑜𝑛𝑑𝑖𝑛𝑔 𝑎𝑢𝑡ℎ𝑜𝑟 𝑚𝑎𝑖𝑙 𝑖𝑑: [email protected] Page 10 of 17

 Frequency Constrained OPF Incorporating UPFC Controller Using DTBO

Table 6 Table 7
Solution of the OPF for variable load with frequency security Solution of the OPF for variable load with frequency security
restriction for the IEEE 30 bus test system (Case 4) restriction for the IEEE 30 bus test system incorporating UPFC
Control (Case 7)
Min. Max. 50% 60% 70% 80% 90% 100%
Parameters Control
PTG1(MW) 50 20073.25 115.95 112.21 136.36 174.48 197.63 Min. Max. 50% 60% 70% 80% 90% 100%
Parameters
PTG2(MW) 20 80 20.97 20 32.44 24.18 26.3 29.03 PTG1(MW) 50 200 75.71 103.87 134.16 135.4 189.27 199.93
PTG5(MW) 15 50 15.05 15 19.27 23.57 18.68 18.45 PTG2(MW) 20 80 20.28 20.71 21.23 47.6 26.92 33.39
PTG8(MW) 10 35 10.06 10 10.83 16.71 20.08 19.52 PTG5(MW) 15 50 15.17 15.85 15.3 16.48 15.46 20.24
PTG11(MW) 10 30 12.31 10 12.64 12.79 11.91 17.51 PTG8(MW) 10 35 10.24 10.46 10.68 11.17 10.28 13.44
PTG13(MW) 12 40 12 2.95 15.4 18.79 12.01 12.16 PTG11(MW) 10 30 10.2 10.24 10.03 10.35 10.15 12.91
𝑉1 (pu) 0.95 1.11.0153 1.0209 1.0444 1.0947 1.0744 1.0704 PTG13(MW) 12 40 12.01 12.02 12.13 12.11 12.29 13.4
𝑉2 (pu) 0.95 1.11.0022 1.0147 1.0395 1.0697 1.0781 1.0423 𝑉1 (𝑝𝑢) 0.95 1.1 1.0145 1.0527 1.0232 1.0199 1.0603 1.1
𝑉5 (pu) 0.95 1.10.9903 1.003 1.0036 1.0533 1.026 1.0227 𝑉2 (𝑝𝑢) 0.95 1.1 1.0037 1.0394 1.0053 1.003 1.0388 1.0828
𝑉8 (pu) 0.95 1.11.0019 1.009 1.0295 1.0242 1.0461 1.0221 𝑉5 (𝑝𝑢) 0.95 1.1 0.9901 1.023 0.9851 0.9823 1.0112 1.051
𝑉8 (𝑝𝑢) 0.95 1.1 0.9939 1.0258 0.9928 0.9786 1.0195 1.0636
𝑉11 (pu) 0.95 1.11.0274 1.0649 1.0231 1.0568 1.0817 1.0548
𝑉11 (𝑝𝑢) 0.95 1.1 1.0399 1.003 1.0806 1.0372 1.0788 1.0641
𝑉13 (pu) 0.95 1.10.9768 1.0088 1.0362 1.0086 0.9701 1.0223
𝑉13 (pu) 0.95 1.1 1.0617 1.0602 1.0417 1.0255 1.064 1.0739
𝑇11 (pu) 0.9 1.11.0547 1.0924 1.084 1.0042 1.0766 1.0523 𝑇11 (pu) 0.9 1.1 0.9659 0.9621 0.9367 1.056 0.9931 1.0686
𝑇12 (pu) 0.9 1.11.0392 0.9116 0.9558 0.9143 0.9076 0.9439 𝑇12 (pu) 0.9 1.1 0.954 0.9876 1.0383 0.9356 1.024 0.962
𝑇15 (pu) 0.9 1.10.9863 0.964 0.949 0.9474 0.9179 0.9414 𝑇15 (𝑝𝑢) 0.9 1.1 0.968 1.0238 0.9746 1.0312 1.0228 1.0532
𝑇36 (pu) 0.9 1.11.0121 0.97 0.9913 0.9511 1.0398 1.0578 𝑇36 (𝑝𝑢) 0.9 1.1 0.9454 0.9775 0.9399 0.9966 0.9901 1.0211
𝑄𝐶10 (MVAR) 0 0.05
0.0166 0.0137 0.0237 0.0488 0.037 0.0347 𝑄𝐶10 (MVAR) 0 0.05 0.0497 0.0488 0.0408 0.0033 0.045 0.0436
𝑄𝐶12 (MVAR) 0 0.05
0.0271 0.045 0.0466 0.02 0.0489 0.0349 𝑄𝐶12 (MVAR) 0 0.05 0.0004 0.0468 0.0414 0.0033 0.041 0.0465
𝑄𝐶15 (MVAR) 0 0.05
0.0024 0.0328 0.0449 0.0323 0.0428 0.0213 𝑄𝐶15 (MVAR) 0 0.05 0.0149 0.0418 0.0305 0.0376 0.0424 0.0365
𝑄𝐶17 (MVAR) 0 0.05
0.0363 0.0292 0.0279 0.0346 0.0128 0.0379 𝑄𝐶17 (MVAR) 0 0.05 0.0314 0.0466 0.0164 0.0258 0.0497 0.0444
𝑄𝐶20 (MVAR) 0 0.05
0.0201 0.0264 0.0303 0.0047 0.0301 0.0302 𝑄𝐶20 (MVAR) 0 0.05 0.0082 0.0268 0.029 0.0263 0.0356 0.0482
𝑄𝐶21 (MVAR) 0 0.05
0.0436 0.0445 0.0224 0.0423 0.0077 0.0455 𝑄𝐶21 (MVAR) 0 0.05 0.0492 0.0489 0.0469 0.0499 0.0499 0.049
𝑄𝐶23 (MVAR) 0 0.05
0.0064 0.0136 0.0238 0.0182 0.0081 0.0159 𝑄𝐶23 (MVAR) 0 0.05 0.0078 0.0109 0.0044 0.0209 0.0217 0.0273
𝑄𝐶24 (MVAR) 0 0.05 0.0359 0.0496 0.0472 0.0496 0.0498 0.0427
𝑄𝐶24 (MVAR) 0 0.05
0.0165 0.0438 0.0473 0.0429 0.0317 0.0311
𝑄𝐶29 (MVAR) 0 0.05 0.0093 0.0178 0.0124 0.017 0.0111 0.026
𝑄𝐶29 (MVAR) 0 0.05
0.0217 0.0096 0.031 0.0069 0.0176 0.0332
Optimal location
Total generation 143.64 173.9 202.79 232.4 263.46 294.3 18-19 18-19 25-27 27-29 29-30 19-20
of UPFC
Total generation cost
($/h) 369.5837 470.419 546.3339 634.5743 739.3651 843.6123 Series source volt-
Emission((t/h) 0.2319 0.262 0.2505 0.2745 0.3336 0.3739 0.001 0.2 0.00133 0.0087 0.0431 0.00213 0.0321 0.0087
age of UPFC (pu)
𝑃𝑙𝑜𝑠𝑠 (MW) 1.93 3.43 4.39 5.67 8.23 10.78 Series source
VD(pu) 0.4946 0.3145 0.6586 1.0588 0.4896 0.6262 phase angle of 0 2𝜋 0.08976 0.7861 0.9871 0.87612 0.6754 0.7861
Frequency deviation with load UPFC (deg)
% max load OS SS(Sec) Shunt source volt-
0.9 1.1 0.9851 1.0011 0.91231 1.0123 1.0043 1.0011
100 0.009316 13.27 age of UPFC (pu)
Frequency
Deviation

90 0.009713 10.02 Shunt source

phase angle of 0 2𝜋 0.0898 1.0023 0.0067 0.6789 0.872 1.0023
80 0.009475 5.46
UPFC (deg.)
70 0.009650 5.20
Total generation 143.61 173.15 203.53 233.11 264.37 293.31
60 0.009521 5.42
Total generation cost ($/h) 367.3087 449.1413 521.0111 622.9392 732.0901 833.3572
50 0.00985 5.00 Emission (t/h) 0.2339 0.2507 0.2807 0.2758 0.3638 0.3792
𝑃𝑙𝑜𝑠𝑠 (MW) 0.0191 3.11 5.15 6.39 9.3 9.92
VD(pu) 0.7603 0.8848 0.5696 0.5069 0.5157 0.9379
Frequency deviation with load
When the objective is to minimize the generating costs at % max load
100
OS
0.00930
SS(Sec)
13.17
Frequency
Deviation

50% load, it has been observed that emissions will increase 90 0.009706 10.00
80 0.009466 5.37
despite a reduction in generation costs. It has been also ob- 70 0.009647 5.16
60 0.009525 5.37
served that when the goal is to minimise the emission, emis- 50 0.009707 4.96
sions are lowered at the same load, but the generation cost
will increase. Multiple objectives are taken into consider-
ation in order to concurrently minimise the cost of gener-
ating and emissions. However in case 3, total generation
cost is 362.8372 ($/h), emission is 0.2462 and frequency de-
viation is (OS=0.01094 and SS=6.87Sec) which displayed
in Table 11 if frequency based security constraints are ig-
nored. For case 6, when frequency based security constraints
are taken into account, at the same load the total genera-
tion cost is 371.6036 ($/h) which is greater as compare to
the case 3 but emission 0.2232 (t/h) and frequency deviation
(OS=0.01078 and SS=6.029Sec) are reduced, as shown in
Table 12. However in case 9, when incorporate UPFC, gen-
Figure 5: IEEE-30 bus system generator data
eration cost 361.805 ($/h) and emission 0.2229 (t/h) both
are reduced which is very much less than case 3 and case
and Fig. 20 shows the frequency deviation at various load
6 and also reduced the frequency deviation (OS=0.010779
(50% to 100%) considering frequency based security con-
and SS=5.978) that is displayed in Table 13.
straint. Fig 9 to Fig 18 depicts the voltage deviation at vari-
Same thing happen for 60%, 70%, 80%, 90% and 100%
ous load (50% to 100%) with frequency based security con-
of load.
straint and incorporating UPFC with frequency based secu-
why only BBO and GWO why not others?
rity constraint. Fig. 6 to Fig 8 shows the convergence graph
Fig. 19 shows the frequency deviation at various load for all load by using DTBO algorithm.
(50% to 100%) without frequency based security constraint

𝐶𝑜𝑟𝑟𝑒𝑠𝑝𝑜𝑛𝑑𝑖𝑛𝑔 𝑎𝑢𝑡ℎ𝑜𝑟 𝑚𝑎𝑖𝑙 𝑖𝑑: [email protected] Page 11 of 17

 Frequency Constrained OPF Incorporating UPFC Controller Using DTBO

Table 9
Table 8
Solution of the OPF for variable load with frequency security
Solution of the OPF for variable load without the frequency
restriction for the IEEE 30 bus test system (Case 5)
security restriction for the IEEE 30 bus test system (Case 2)
Control
Control Min. Max. 50% 60% 70% 80% 90% 100%
Min. Max. 50% 60% 70% 80% 90% 100% Parameters
Parameters PTG1(MW) 50 200 50 51.66 50 50 50.09 59.63
PTG1(MW) 0 200 42.74 52.82 73.78 73.68 102.51 132.46 PTG2(MW) 20 80 24.32 31.24 39.51 42.36 58.92 73.1
PTG2(MW) 0 80 53.34 70.38 79.34 78.37 79.45 79.73 PTG5(MW) 15 50 24.87 34.89 39.78 47.63 48.75 49.81
PTG5(MW) 0 50 49.72 49.13 48.56 48.59 49.57 48.57 PTG8(MW) 10 35 13.91 17.2 21.67 29.89 31.63 34.86
PTG8(MW) 0 35 0 0 0 9.82 9.94 9.9 PTG11(MW) 10 30 20.92 25.6 27.26 28.39 29.99 29.89
PTG11(MW) 0 30 0 0 0 9.83 9.09 9.9 PTG13(MW) 12 40 25.85 29.29 32.7 39.87 37.96 39.47
PTG13(MW) 0 40 0 0 0 9.85 9.91 9.81 𝑉1 (pu) 0.95 1.1 0.9525 0.9519 0.9665 0.9749 1.0771 1.0121
𝑉1 (pu) 0.95 1.1 0.9872 1.0304 1.0671 1.0416 1.0241 1.0628 𝑉2 (pu) 0.95 1.1 1.0982 1.0984 1.0855 1.0784 1.0686 1.0088
𝑉2 (pu) 0.95 1.1 0.9889 1.0284 1.0624 1.0349 1.0106 1.0507 𝑉5 (pu) 0.95 1.1 0.9937 0.9569 1.0185 1.0346 1.0501 0.994
𝑉5 (pu) 0.95 1.1 1.0223 1.019 1.0509 1.0205 0.9737 1.0314 𝑉8 (pu) 0.95 1.1 1.0606 1.0341 1.0166 0.9893 1.0601 0.999
𝑉8 (pu) 0.95 1.1 1.0563 1.0082 1.0422 1.0164 0.9793 1.0217 𝑉11 (pu) 0.95 1.1 1.0878 1.0418 1.0147 0.9896 1.0772 1.097
𝑉11 (pu) 0.95 1.1 0.9616 1.0721 1.0878 1.095 1.0812 1.0981 𝑉13 (pu) 0.95 1.1 0.9654 1.0291 1.0503 0.9828 1.0835 1.0981
𝑉13 (pu) 0.95 1.1 0.9542 1.0855 1.0842 1.0928 1.0719 1.0987 𝑇11 (pu) 0.9 1.1 0.9726 1.0736 1.0703 1.008 1.04 0.9297
𝑇11 (pu) 0.9 1.1 1.007 0.9534 0.9796 0.9621 0.958 0.9849 𝑇12 (pu) 0.9 1.1 1.0834 0.9451 0.9019 1.0125 0.9478 0.9
𝑇12 (pu) 0.9 1.1 1.0055 0.9488 1.0029 0.9477 0.9409 0.9075 𝑇15 (pu) 0.9 1.1 0.9384 1.0237 1.0001 1.0922 1.0286 0.9195
𝑇15 (pu) 0.9 1.1 1.024 0.9566 1.004 0.9939 0.9825 0.9673 𝑇36 (pu) 0.9 1.1 1.0131 1.0037 0.9548 0.9128 0.9888 0.9
𝑇36 (pu) 0.9 1.1 0.9524 0.944 0.9717 0.9387 0.9433 0.9354 𝑄𝐶10 (MVAR) 0 5 1.95 0.0203 0.0448 0.0247 0.0368 0.0491
𝑄𝐶10 (MVAR) 0 5 0.17 0.0496 0.0483 0.0152 0.0459 0.0466 𝑄𝐶12 (MVAR) 0 5 0.42 0.0493 0.0247 0.02 0.0434 0.0497
𝑄𝐶12 (MVAR) 0 5 3.69 0.0499 0.0497 0.0323 0.0232 0.0465 𝑄𝐶15 (MVAR) 0 5 2.11 0.0467 0.0236 0.0153 0.0417 0.0485
𝑄𝐶15 (MVAR) 0 5 1.7 0.0286 0.029 0.0307 0.0365 0.0473 𝑄𝐶17 (MVAR) 0 5 0.0137 0.0417 0.0185 0.0379 0.0487 0.0494
𝑄𝐶17 (MVAR) 0 5 0.09 0.0418 0.0425 0.0478 0.0475 0.0484 𝑄𝐶20 (MVAR) 0 5 0.0122 0.017 0.0059 0.0363 0.028 0.0482
𝑄𝐶20 (MVAR) 0 5 4.27 0.0224 0.0314 0.028 0.0371 0.0425 𝑄𝐶21 (MVAR) 0 5 0.0253 0.0447 0.0256 0.0217 0.0428 0.0493
𝑄𝐶21 (MVAR) 0 5 3.42 0.0485 0.0479 0.0422 0.04 0.0498 𝑄𝐶23 (MVAR) 0 5 0.0172 0.0138 0.0047 0.047 0.0256 0.0295
𝑄𝐶23 (MVAR) 0 5 0.53 0.0074 0.0126 0.0173 0.0377 0.0322 𝑄𝐶24 (MVAR) 0 5 0.0333 0.0458 0.0492 0.0482 0.0487 0.0498
𝑄𝐶24 (MVAR) 0 5 0.51 0.0437 0.0457 0.0453 0.0476 0.0464 𝑄𝐶29 (MVAR) 0 5 0.0226 0.0135 0.0019 0.0081 0.0329 0.027
𝑄𝐶29 (MVAR) 0 5 1.08 0.0185 0.02 0.0153 0.02 0.0261 Total generation 159.87 189.88 210.92 238.14 257.34 286.76
Total generation 145.8 172.33 201.68 230.14 260.47 290.37 Total generation cost
($/h) 463.7147 595.1809 691.3339 829.8473 905.2974 1018.7865
Total generation cost
($∕ℎ) 484.8687 562.7763 635.1823 743.6723 832.7961 922.6314 Emission((t/h) 0.213 0.2058 0.2012 0.1975 0.1967 0.1998
Emission((t/h) 0.2269 0.2292 0.237 0.2244 0.2419 0.2708 𝑃𝑙𝑜𝑠𝑠 (MW) 18.17 19.84 12.54 11.43 2.28 3.35
𝑃𝑙𝑜𝑠𝑠 (MW) 4.1 2.29 3.3 3.42 5.42 6.95 VD(pu) 0.5969 0.4295 0.4957 0.5721 1.2661 1.5282
VD(pu) 0.5234 1.3238 1.4163 1.2347 0.672 1.4739 Frequency deviation with load
Frequency deviation with load % max load OS SS(Sec)
% max load OS SS(Sec) 100 0.009399 17.012
Frequency
Deviation

100 0.009613 17.1

Frequency
Deviation

90 0.009712 10.020
90 0.009731 10.05 80 0.009713 8.091
80 0.009725 8.33 70 0.0098397 8.622
70 0.009856 8.73 60 0.009790 7.02
60 0.009892 7.56 50 0.010045 5.806
50 0.01082 6.41

Figure 6: Convergence of DTBO algorithm for various load

of IEEE-30 bus system without frequency based security con- Figure 7: Convergence of DTBO algorithm for various load of
straint. IEEE-30 bus system with frequency based security constraint.

𝐶𝑜𝑟𝑟𝑒𝑠𝑝𝑜𝑛𝑑𝑖𝑛𝑔 𝑎𝑢𝑡ℎ𝑜𝑟 𝑚𝑎𝑖𝑙 𝑖𝑑: [email protected] Page 12 of 17

 Frequency Constrained OPF Incorporating UPFC Controller Using DTBO

Table 10
Table 11
Solution of the OPF for variable load with frequency security
Solution of the OPF for variable load without the frequency
restriction for the IEEE 30 bus test system incorporating UPFC
security restriction for the IEEE 30 bus test system (Case 3)
(Case 8)
Control
Control Min. Max. 50% 60% 70% 80% 90% 100%
Min. Max. 50% 60% 70% 80% 90% 100% Parameters
Parameters PTG1(MW) 0 200 84.35 85.13 99.28 84.99 87.5 98.9
PTG1(MW) 50 200 50 50 50 50 50 60.34 PTG2(MW) 0 80 46.36 70.55 77.29 66.96 71.93 77.53
PTG2(MW) 20 80 22.69 35.58 43.01 43.98 59 72.54 PTG5(MW) 0 50 13.67 17.7 26.16 17.4 24.64 27.4
PTG5(MW) 15 50 28.73 36.88 39.13 49.04 48.56 49.29
PTG8(MW) 0 35 0 0 0 33.8 33.9 35
PTG8(MW) 10 35 15.86 14.27 18.49 31.47 32.78 34.97
PTG11(MW) 0 30 0 0 0 18.17 21.15 25.77
PTG11(MW) 10 30 20.59 22.59 28.73 29.81 29.57 29.94
PTG13(MW) 0 40 0 0 0 9 20.13 24.25
PTG13(MW) 12 40 21.45 29.19 36.28 38.33 38.99 39.54
𝑉1 (pu) 0.95 1.1 0.9659 0.9529 0.9642 1.067 1.0125 1.0454 𝑉1 (pu) 0.95 1.1 1.0499 1.0856 1.084 1.0769 1.0578 1.0332
𝑉2 (pu) 0.95 1.1 1.0976 1.0841 1.0975 0.9576 0.9926 1.0423 𝑉2 (pu) 0.95 1.1 1.0424 1.0791 1.0712 1.0677 1.0478 1.0226
𝑉5 (pu) 0.95 1.1 1.0807 0.961 1.0386 1.0818 0.9934 1.0256 𝑉5 (pu) 0.95 1.1 1.0276 1.0606 1.0539 1.0422 1.0237 0.9948
𝑉8 (pu) 0.95 1.1 0.9597 1.0828 1.0173 0.9783 0.9779 1.0304 𝑉8 (pu) 0.95 1.1 1.0276 1.0605 1.0523 1.0512 1.0363 1.002
𝑉11 (pu) 0.95 1.1 1.0989 1.0181 1.0445 1.0506 1.0672 1.0953 𝑉11 (pu) 0.95 1.1 1.0618 1.0537 1.0963 1.0987 1.0994 1.0988
𝑉13 (pu) 0.95 1.1 1.0968 0.9738 1.0204 0.9907 0.9712 1.0792 𝑉13 (pu) 0.95 1.1 1.0786 1.0639 1.097 1.098 1.0979 1.099
𝑇11 (pu) 0.9 1.1 0.9983 0.9188 1.0344 1.0276 0.9938 1.026 𝑇11 (pu) 0.9 1.1 1.0356 1.0734 1.059 1.0257 0.9933 0.9455
𝑇12 (pu) 0.9 1.1 0.9035 0.9412 0.956 1.0514 0.9773 0.9127 𝑇12 (pu) 0.9 1.1 0.9 0.9456 0.9 0.9 0.9 0.9
𝑇15 (pu) 0.9 1.1 0.9735 0.9879 0.9251 1.0963 0.9786 0.9801 𝑇15 (pu) 0.9 1.1 0.9784 1.0194 0.9875 0.964 0.954 0.9337
𝑇36 (pu) 0.9 1.1 0.9157 0.9123 1.0807 0.9373 1.0408 0.9578 𝑇36 (pu) 0.9 1.1 0.9611 1.0076 0.9668 0.9493 0.9371 0.9107
𝑄𝐶10 (MVAR) 0 0.05 0.05 0.0475 0.0465 0.0141 0.0365 0.0489 𝑄𝐶10 (MVAR) 0 5 0.0461 0.0485 0.0363 0.05 0.0485 0.0477
𝑄𝐶12 (MVAR) 0 0.05 0.0494 0.0275 0.0206 0.0073 0.0058 0.0483 𝑄𝐶12 (MVAR) 0 5 0.049 0.05 0.0493 0.0498 0.0481 0.0497
𝑄𝐶15 (MVAR) 0 0.05 0.0206 0.0023 0.046 0.0309 0.0118 0.041 𝑄𝐶15 (MVAR) 0 5 0.0191 0.0245 0.0282 0.0396 0.0442 0.0468
𝑄𝐶17 (MVAR) 0 0.05 0.0342 0.0419 0.0019 0.0489 0.0418 0.048 𝑄𝐶17 (MVAR) 0 5 0.0322 0.0393 0.0455 0.0489 0.0494 0.0482
𝑄𝐶20 (MVAR) 0 0.05 0.0201 0.0112 0.0149 0.042 0.0274 0.0436 𝑄𝐶20 (MVAR) 0 5 0.0193 0.0248 0.0253 0.0313 0.0372 0.0463
𝑄𝐶21 (MVAR) 0 0.05 0.0495 0.0365 0.0321 0.0182 0.0053 0.0475
𝑄𝐶21 (MVAR) 0 5 0.0495 0.0494 0.0491 0.0498 0.05 0.0488
𝑄𝐶23 (MVAR) 0 0.05 0.008 0.0141 0.0239 0.0472 0.0114 0.0429
𝑄𝐶23 (MVAR) 0 5 0.0068 0.0102 0.0113 0.0158 0.0234 0.0314
𝑄𝐶24 (MVAR) 0 0.05 0.0337 0.0281 0.0143 0.0258 0.0202 0.0464
𝑄𝐶29 (MVAR) 0 0.05 0.0096 0.0271 0.0381 0.0331 0.0357 0.0149
𝑄𝐶24 (MVAR) 0 5 0.035 0.0459 0.0483 0.0492 0.05 0.05
Optimal location 𝑄𝐶29 (MVAR) 0 5 0.012 0.0176 0.0193 0.0185 0.0238 0.027
18-19 18-19 25-27 27-29 29-30 19-20 Total generation 144.38 173.3712 202.73 230.3198 259.2484 288.849
of UPFC
Series source volt- Total generation cost ($∕ℎ) 362.8372 461.6157 568.7462 684.072 799.3587 914.2618
0.001 0.2 0.00163 0.0094 0.0456 0.00342 0.0367 0.0039 Emission((t/h) 0.2462 0.2465 0.2546 0.2274 0.222 0.2273
age of UPFC (pu)
Series source 𝑃𝑙𝑜𝑠𝑠 (MW) 2.68 3.33 4.35 3.6 4.19 5.45
phase angle of 0 2𝜋 0.08778 0.7561 0.8791 0.85612 0.6984 0.8761 VD(pu) 1.3349 1.2222 1.7381 1.8994 1.7259 1.4781
UPFC (deg) Frequency deviation with load
Shunt source volt- % max load OS SS(Sec)
0.9 1.1 0.9341 1.0013 0.92371 1.0783 1.0043 1.0011
age of UPFC (pu) 100 0.009449 17.18
Frequency
Deviation

Shunt source 90 0.009712 10.03

phase angle of 0 2𝜋 0.0898 1.0023 0.0067 0.6889 0.972 1.0033 80 0.009736 8.37
UPFC (deg.) 70 0.009877 8.628
Total generation 159.32 188.51 215.64 242.63 258.9 286.62 60 0.009887 7.76
Total generation cost ($/h) 463.0364 590.4484 712.1406 851.193 911.9259 1015.9593
50 0.01094 6.87
Emission (t/h) 0.2127 0.20574 0.2003 0.197 0.1965 0.1997
𝑃𝑙𝑜𝑠𝑠 (MW) 0.1762 18.43 17.14 15.49 3.69 3.22
VD(pu) 1.4742 1.4394 0.8284 0.3712 0.6106 1.1584
Frequency deviation with load
% max load OS SS(Sec)
100 0.009388 17.08
Frequency
Deviation

90 0.009712 10.02
80 0.0097101 8.07
70 0.009839 8.598
60 0.009782 6.99
50 0.009743 5.77

Figure 9: For 50% load with frequency based security con-

straint.

Figure 8: Convergence of DTBO algorithm for various load of Figure 10: For 50% load with frequency based security con-
IEEE-30 bus system with frequency based security constraint straint and UPFC
and UPFC.

𝐶𝑜𝑟𝑟𝑒𝑠𝑝𝑜𝑛𝑑𝑖𝑛𝑔 𝑎𝑢𝑡ℎ𝑜𝑟 𝑚𝑎𝑖𝑙 𝑖𝑑: [email protected] Page 13 of 17

 Frequency Constrained OPF Incorporating UPFC Controller Using DTBO

Table 12 Table 13
Solution of the OPF for variable load with frequency security Solution of the OPF for variable load with frequency security
restriction for the IEEE 30 bus test system (case 6) restriction for the IEEE 30 bus test system incorporating UPFC
Control
Min. Max. 50% 60% 70% 80% 90% 100% (Case 9)
Parameters
PTG1(MW) 50 20057.78 81.8 114.1 89.08 103.48 108.86 Control
Min. Max. 50% 60% 70% 80% 90% 100%
PTG2(MW) 20 80 35.97 40 38.25 53.59 56.55 64.24 Parameters
PTG5(MW) 15 50 17.56 18.71 18.48 22.91 24.38 26.93 PTG1(MW) 50 200 57.34 81.6 110.91 93.27 107.21 115.4
PTG8(MW) 10 35 10 10 10 26.59 34.24 34.99 PTG2(MW) 20 80 36.07 39.68 39.12 52.98 56.11 60.93
PTG11(MW) 10 30 10 10 10 18.54 20.73 27.28 PTG5(MW) 15 50 17.42 18.92 18.86 23.03 23.96 26.39
PTG8(MW) 10 35 10.16 10.07 10.03 24.41 33.09 34.43
PTG13(MW) 12 40 12 12 12 19.44 20.08 26.14
PTG11(MW) 10 30 10.11 10.1 11.36 18.66 20.2 27.23
𝑉1 (𝑝𝑢) 0.95 1.10.9979 1.0685 1.0149 1.0758 1.0763 1.0971 PTG13(MW) 12 40 12.17 12.21 12.03 18.05 19.26 24.56
𝑉2 (𝑝𝑢) 0.95 1.10.9926 1.062 1.0063 1.0688 1.0642 1.0889 𝑉1 (𝑝𝑢) 0.95 1.1 1.0271 1.0543 1.0778 1.0759 1.0744 1.0889
𝑉5 (𝑝𝑢) 0.95 1.10.9785 1.0449 0.9814 1.0451 1.0392 1.0661 𝑉2 (𝑝𝑢) 0.95 1.1 1.0182 1.0437 1.0649 1.0651 1.0596 1.0756
𝑉8 (𝑝𝑢) 0.95 1.10.9816 1.0459 0.9819 1.0564 1.0509 1.0775 𝑉5 (𝑝𝑢) 0.95 1.1 0.9994 1.0235 1.0398 1.0472 1.0393 1.0464
𝑉11 (𝑝𝑢) 0.95 1.11.031 1.0363 1.0878 1.0629 1.0985 1.0984 𝑉8 (𝑝𝑢) 0.95 1.1 1.0064 1.0259 1.0495 1.0537 1.0507 1.0579
𝑉13 (𝑝𝑢) 0.95 1.11.0467 1.0469 1.0898 1.0906 1.0989 1.0999 𝑉11 (𝑝𝑢) 0.95 1.1 1.0595 1.0512 1.03 0.9962 1.0589 1.0992
𝑇11 (𝑝𝑢) 0.9 1.10.9493 1.044 0.937 1.0248 0.9777 1.0242 𝑉13 (𝑝𝑢) 0.95 1.1 1.0373 1.0584 1.0271 1.0387 1.0068 1.0117
𝑇12 (𝑝𝑢) 0.9 1.10.9664 0.9747 0.9 0.9546 0.9529 0.9409 𝑇11 (𝑝𝑢) 0.9 1.1 0.9481 0.9745 1.0733 1.0731 1.0618 1.0549
𝑇15 (𝑝𝑢) 0.9 1.10.9623 1.0283 0.9266 1.0312 0.9758 0.9943 𝑇12 (𝑝𝑢) 0.9 1.1 1.0998 1.0207 1 0.9723 1.0911 0.9999
𝑇36 (𝑝𝑢) 0.9 1.10.9425 1.0084 0.9055 0.9802 0.9515 0.9711 𝑇15 (𝑝𝑢) 0.9 1.1 1.009 1.0102 1.0531 1.095 1.063 0.9793
𝑄𝐶10 (MVAR) 0 5 0.0488 0.0426 0.0496 0.0497 0.0485 0.0498 𝑇36 (𝑝𝑢) 0.9 1.1 0.9871 0.9746 1.0237 1.0387 1.0479 0.984
𝑄𝐶10 (MVAR) 0 5 1.83 3.4 1.83 4.37 0.99 4.59
𝑄𝐶12 (MVAR) 0 5 0.0499 0.0479 0.0491 0.0483 0.0497 0.0498
𝑄𝐶12 (MVAR) 0 5 0.21 4.95 2.98 4.91 2.39 4.83
𝑄𝐶15 (MVAR) 0 5 0.0204 0.0266 0.0247 0.0332 0.0455 0.0492
𝑄𝐶15 (MVAR) 0 5 2.29 2.13 2.18 2.58 3.67 4.94
𝑄𝐶17 (MVAR) 0 5 0.0346 0.0356 0.0463 0.0489 0.0483 0.049 𝑄𝐶17 (MVAR) 0 5 1.94 4.68 4.46 4.16 4.81 4.87
𝑄𝐶20 (MVAR) 0 5 0.0198 0.0265 0.0275 0.0361 0.0392 0.0458 𝑄𝐶20 (MVAR) 0 5 1.57 1.01 3.42 2.91 1.81 4.95
𝑄𝐶21 (MVAR) 0 5 0.05 0.0486 0.0492 0.0499 0.0486 0.0497 𝑄𝐶21 (MVAR) 0 5 4.37 4.86 4.47 4.56 4.95 4.83
𝑄𝐶23 (MVAR) 0 5 0.0078 0.0095 0.0144 0.0198 0.0232 0.0295 𝑄𝐶23 (MVAR) 0 5 1.01 0.93 1.61 1.3 4.54 4.44
𝑄𝐶24 (MVAR) 0 5 0.0398 0.0476 0.0495 0.0456 0.0489 0.0489 𝑄𝐶24 (MVAR) 0 5 3.03 4.51 4.64 4.93 4.44 4.63
𝑄𝐶29 (MVAR) 0 5 0.0108 0.015 0.0206 0.0173 0.0224 0.0272 𝑄𝐶29 (MVAR) 0 5 1.22 1.41 1.56 2.33 3.26 2.37
Total generation 143.3103 172.516 202.8348 230.1571 259.4586 288.4308 Optimal location
18-19 18-19 25-27 27-29 29-30 19-20
Total generation cost
($∕ℎ) 371.6036 462.4484 539.5325 686.2913 785.9639 902.8624 of UPFC
Emission((t/h) 0.2232 0.2305 0.254 0.2232 0.2313 0.2318 Series source volt-
0.001 0.2 0.00163 0.0094 0.0456 0.00342 0.0367 0.0039
𝑃𝑙𝑜𝑠𝑠 (MW) 1.61 2.48 4.45 3.44 4.4 5.03 age of UPFC (pu)
VD(pu) 0.6803 0.81 1.4746 1.3552 1.7755 1.8962 Shunt source
phase angle of 0 2𝜋 0.08778 0.7561 0.8791 0.85612 0.6984 0.8761
Frequency deviation with load
UPFC (deg.)
% max load OS SS(Sec) Shunt source volt-
100 0.009397 16.09 0.9 1.1 0.9341 1.0013 0.92371 1.0783 1.0043 1.0011
Frequency
Deviation

age of UPFC (pu)

90 0.009698 9.87 Shunt source
80 0.009609 8.124 phase angle of 0 2𝜋 0.0898 1.0023 0.0067 0.6889 0.972 1.0033
70 0.0098321 8.509 UPFC (deg.)
60 0.009724 7.18 Total generation 143.2542 172.574 202.3004 230.3929 259.8258 288.94
50 0.01078 6.029 Total generation cost ($/h) 361.805 461.0591 542.2149 681.227 780.8432 892.9907
Emission (t/h) 0.2229 0.2302 0.2503 0.2225 0.2309 0.2308
𝑃𝑙𝑜𝑠𝑠 (MW) 0.0155 2.53 3.92 3.67 4.76 5.53
VD(pu) 0.2837 0.8671 0.3688 0.3684 0.6035 0.9067
Frequency deviation with load
% max load OS SS(Sec)
100 0.00939689 16.088
Frequency
Deviation

90 0.0096898 9.869
80 0.0096089 8.098
70 0.00983189 8.5088
60 0.0097089 6.87
50 0.010779 5.978

Table 14
Comparison of the fuel cost ($/h) for six load scenarios with-
out the frequency security restriction for the IEEE 30 bus test
Figure 11: For 60% load with frequency based security con- system using different algorithm
straint
Algorithm 50% 60% 70% 80% 90% 100%
DTBO 368.9182 458.4223 541.5061 633.8415 738.8719 843.5678
BBO 370.4632 460.5637 542.7865 635.8976 740.3443 844.9088
GWO 389.7831 459.0981 542.0973 634.4555 739.0032 844.0943

Table 15
Comparison of the fuel cost ($/h) for six load scenarios with
frequency security restriction for the IEEE 30 bus test system
using different algorithm
Algorithm 50% 60% 70% 80% 90% 100%
DTBO 369.5837 470.419 546.3339 634.5743 739.3651 843.6123
BBO 371.8911 472.7651 548.9834 636.7763 740.8953 844.9833
Figure 12: For 60% load with frequency based security con- GWO 370.4454 471.4322 547.542 635.0983 740.1102 843.9987

straint and UPFC

𝐶𝑜𝑟𝑟𝑒𝑠𝑝𝑜𝑛𝑑𝑖𝑛𝑔 𝑎𝑢𝑡ℎ𝑜𝑟 𝑚𝑎𝑖𝑙 𝑖𝑑: [email protected] Page 14 of 17

 Frequency Constrained OPF Incorporating UPFC Controller Using DTBO

Table 16
Comparison of the fuel cost ($/h) for six load scenarios with
frequency security restriction and UPFC for the IEEE 30 bus
test system using different algorithm
Algorithm 50% 60% 70% 80% 90% 100%
DTBO 367.3087 449.1413 521.0111 622.9392 732.0901 833.3572
BBO 368.8879 450.8792 522.5562 624.8223 734.6322 834.7865
GWO 368.0123 450.1015 521.9827 623.7332 733.5221 833.9876

Table 17
Comparison of the emission (t/h) for six load scenarios with-
out the frequency security restriction for the IEEE 30 bus test
system using different algorithm Figure 15: For 80% load with frequency based security con-
straint
Algorithm 50% 60% 70% 80% 90% 100%
DTBO 0.2269 0.2292 0.2372 0.2244 0.2419 0.2708
BBO 0.2367 0.2356 0.2452 0.2354 0.2499 0.2812
GWO 0.2288 0.2312 0.2411 0.2279 0.2454 0.2712

Table 18
Comparison of the emission (t/h) for six load scenarios with
frequency security restriction for the IEEE 30 bus test system
using different algorithm
Algorithm 50% 60% 70% 80% 90% 100%
DTBO 0.213 0.2058 0.2012 0.1975 0.1967 0.1998
BBO 0.2254 0.2133 0.2145 0.2189 0.2154 0.2123
GWO 0.2211 0.2106 0.2099 0.2012 0.2056 0.2077
Figure 16: For 80% load with frequency based security con-
Table 19 straint and UPFC
Comparison of the emission (t/h) for six load scenarios with
frequency security restriction and UPFC for the IEEE 30 bus
test system using different algorithm
Algorithm 50% 60% 70% 80% 90% 100%
DTBO 0.2127 0.2057 0.2003 0.197 0.1965 0.1997
BBO 0.2254 0.2154 0.2113 0.2078 0.2008 0.2045
GWO 0.2156 0.2087 0.2045 0.2055 0.1997 0.2011

Figure 17: For 90% load with frequency based security con-
straint

Figure 13: For 70% load with frequency based security con-
straint

Figure 18: For 100% load with frequency based security con-
Figure 14: For 70% load with frequency based security con- straint
straint and UPFC

𝐶𝑜𝑟𝑟𝑒𝑠𝑝𝑜𝑛𝑑𝑖𝑛𝑔 𝑎𝑢𝑡ℎ𝑜𝑟 𝑚𝑎𝑖𝑙 𝑖𝑑: [email protected] Page 15 of 17

 Frequency Constrained OPF Incorporating UPFC Controller Using DTBO

UPFC with frequency based security constraint, the more

reduction in the values of the aforesaid OPFs are observed
and frequency deviation are same as in the second case.
Therefore, for minimize generation cost and improving
frequency deviation at any load the frequency based se-
curity constraint must be introduced by incorporating UPFC.

Declaration
Ethical approval
There are no human subject in this article and informed
consent is not applicable.
Figure 19: Frequency deviation at various load without fre-
quency based security constraint Consent for publication
Not applicable.

Conflicts of Interest
The authors declare no conflict of interest.

Funding
The authors received no specific funding for this study
from their institutes.

Data Availability Statement

The datasets generated during and/or analysed during the
current study are available from the corresponding author on
reasonable request.

Authors’ contributions
A.R., S.D., A.B., S.B., R.K.C., M.A.K.A., M.K., S.V.
Figure 20: Frequency deviation at various load with frequency worked on Conceptualizations, methodology, formal analy-
based security constraint
sis, and writing—original draft preparation, while R.A.K.,
S.M. and M.A.S. participated in writing—review and edit-
5.1. Statistical assessment ing, validation, supervision, project administration. All au-
From Table 7, it is observed that the statistical values thors have read and agreed to the published version of the
which are generated from the outcomes, produced by DTBO, manuscript.
are improved with respect to results in Table 5 and Table 6
which leads to conclude that the performance by incorpo-
rating UPFC with frequency constraint is superior than the References
other two cases in OPF problem. [1] Nga Nguyen and Joydeep Mitra. Reliability of power system with high
wind penetration under frequency stability constraint. IEEE Transac-
tions on Power Systems, 33(1):985–994, 2017.
6. Conclusions [2] Nga Nguyen and Joydeep Mitra. An analysis of the effects and de-
pendency of wind power penetration on system frequency regulation.
The current work has employed DTBO to find out best IEEE Transactions on Sustainable Energy, 7(1):354–363, 2015.
possible solution for an OPF problem where IEEE 30-bus [3] Javad Lavei, Anders Rantzer, and Stephen Low. Power flow opti-
network with thermal power sources, is considered as mization using positive quadratic programming. IFAC Proceedings
Volumes, 44(1):10481–10486, 2011.
test system in which frequency based security constraints
[4] Sirote Khunkitti, Apirat Siritaratiwat, and Suttichai Premrudeep-
and UPFC are incorporated. To improve the frequency reechacharn. Multi-objective optimal power flow problems based on
deviation at lower load, frequency based security constraint slime mould algorithm. Sustainability, 13(13):7448, 2021.
are considered. In the first case of this study, it is found that [5] Mehak Kohli and Sankalap Arora. Chaotic grey wolf optimization
without frequency based security constraint significantly algorithm for constrained optimization problems. Journal of compu-
tational design and engineering, 5(4):458–472, 2018.
reduced the generation cost, at lower load but frequency de-
[6] Ali Asghar Heidari and Parham Pahlavani. An efficient modified grey
viation are increases at these load. In the second case of this wolf optimizer with lévy flight for optimization tasks. Applied Soft
study while frequency based security constraint has been Computing, 60:115–134, 2017.
introduced the generation cost is increased at lower load but [7] Seyedali Mirjalili. The ant lion optimizer. Advances in engineering
frequency deviation are reduced and it is almost constant software, 83:80–98, 2015.
at all load. In the third case of this study while incorporate

𝐶𝑜𝑟𝑟𝑒𝑠𝑝𝑜𝑛𝑑𝑖𝑛𝑔 𝑎𝑢𝑡ℎ𝑜𝑟 𝑚𝑎𝑖𝑙 𝑖𝑑: [email protected] Page 16 of 17

 Frequency Constrained OPF Incorporating UPFC Controller Using DTBO

[8] Aboul Ella Hassanien, Rizk M Rizk-Allah, and Mohamed Elhoseny. 25(14):9389–9426, 2021.
A hybrid crow search algorithm based on rough searching scheme [25] Tarek A Boghdady and Youssef A Mohamed. Reactive power com-
for solving engineering optimization problems. Journal of Ambient pensation using statcom in a pv grid connected system with a modified
Intelligence and Humanized Computing, pages 1–25, 2018. mppt method. Ain Shams Engineering Journal, 14(8):102060, 2023.
[9] Partha P Biswas, PN Suganthan, and Gehan AJ Amaratunga. Optimal [26] HREH Bouchekara, MA Abido, AE Chaib, and R Mehasni. Optimal
power flow solutions incorporating stochastic wind and solar power. power flow using the league championship algorithm: A case study
Energy conversion and management, 148:1194–1207, 2017. of the algerian power system. Energy conversion and management,
[10] Ehab E Elattar. Optimal power flow of a power system incorporat- 87:58–70, 2014.
ing stochastic wind power based on modified moth swarm algorithm. [27] Hamed Ahmadi and Hassan Ghasemi. Security-constrained unit com-
IEEE Access, 7:89581–89593, 2019. mitment with linearized system frequency limit constraints. IEEE
[11] Mohammad Zohrul Islam, Noor Izzri Abdul Wahab, Veerapandiyan Transactions on Power Systems, 29(4):1536–1545, 2014.
Veerasamy, Nashiren Farzilah Mailah, Hashim Hizam, Mohamad [28] Philip M Anderson and Mahmood Mirheydar. A low-order system
Nasrun Mohd Nasir, and Arangarajan Vinayagam. Generation fuel frequency response model. IEEE transactions on power systems,
cost and loss minimization using salp swarm algorithm based optimal 5(3):720–729, 1990.
power flow. In 2020 International Conference on Computer Commu- [29] Qing-Chang Zhong and George Weiss. Synchronverters: Inverters
nication and Informatics (ICCCI), pages 1–6. IEEE, 2020. that mimic synchronous generators. IEEE transactions on industrial
[12] AB Layth, AK Murtadha, and A Jaleel. Solving optimal power electronics, 58(4):1259–1267, 2010.
flow problem using improved differential evolution algorithm. Int. [30] Mohammad Dehghani, Eva Trojovská, and Pavel Trojovskỳ. A new
J. Electr. Electron. Eng. Telecommun, 11:146–155, 2022. human-based metaheuristic algorithm for solving optimization prob-
[13] Anbo Meng, Cong Zeng, Peng Wang, De Chen, Tianmin Zhou, Xi- lems on the base of simulation of driving training process. Scientific
aoying Zheng, and Hao Yin. A high-performance crisscross search reports, 12(1):9924, 2022.
based grey wolf optimizer for solving optimal power flow problem. [31] Partha P Biswas, Ponnuthurai N Suganthan, Rammohan Mallipeddi,
Energy, 225:120211, 2021. and Gehan AJ Amaratunga. Optimal reactive power dispatch with
[14] Salma Abd El-sattar, Salah Kamel, Mohamed Ebeed, and Francisco uncertainties in load demand and renewable energy sources adopt-
Jurado. An improved version of salp swarm algorithm for solving ing scenario-based approach. Applied Soft Computing, 75:616–632,
optimal power flow problem. Soft Computing, 25:4027–4052, 2021. 2019.
[15] Mohamed AM Shaheen, Hany M Hasanien, and Ahmed Al-Durra.
Solving of optimal power flow problem including renewable energy
resources using heap optimization algorithm. IEEE Access, 9:35846–
35863, 2021.
[16] Ravi Kumar Avvari and Vinod Kumar DM. A new hybrid evolution-
ary algorithm for multi-objective optimal power flow in an integrated
we, pv, and pev power system. Electric Power Systems Research,
214:108870, 2023.
[17] Mallala Balasubbareddy, Divyanshi Dwivedi, Garikamukkala
Venkata Krishna Murthy, and Sowjan Kumar Kotte. Optimal power
flow solution with current injection model of generalized interline
power flow controller using ameliorated ant lion optimization.
International Journal of Electrical and Computer Engineering,
13(1):1060, 2023.
[18] Inam Ullah Khan, Nadeem Javaid, Kelum AA Gamage, C James Tay-
lor, Sobia Baig, and Xiandong Ma. Heuristic algorithm based optimal
power flow model incorporating stochastic renewable energy sources.
IEEE Access, 8:148622–148643, 2020.
[19] Sirote Khunkitti, Apirat Siritaratiwat, Suttichai Premrudeep-
reechacharn, Rongrit Chatthaworn, and Neville R Watson. A hybrid
da-pso optimization algorithm for multiobjective optimal power flow
problems. Energies, 11(9):2270, 2018.
[20] Douglas Rodrigues, Gustavo Henrique de Rosa, Leandro Aparecido
Passos, and João Paulo Papa. Adaptive improved flower pollination
algorithm for global optimization. Nature-inspired computation in
data mining and machine learning, pages 1–21, 2020.
[21] Pramod Kumar Gouda, Ashwin Kumar Sahoo, and PK Hota. Optimal
power flow including unified power flow controller in a deregulated
environment. International Journal of Applied Engineering Research,
10(1):505–522, 2015.
[22] Amal Amin Mohamed, Salah Kamel, Mohamed H Hassan, and Ab-
dalla A Ibrahim. Northern goshawk optimization algorithm for opti-
mal power flow with facts devices in wind power integrated electri-
cal networks. Electric Power Components and Systems, pages 1–23,
2023.
[23] Ankur Maheshwari, Yog Raj Sood, and Supriya Jaiswal. Investiga-
tion of optimal power flow solution techniques considering stochastic
renewable energy sources: Review and analysis. Wind Engineering,
47(2):464–490, 2023.
[24] Serhat Duman, Jie Li, Lei Wu, and Nuran Yorukeren. Symbiotic
organisms search algorithm-based security-constrained ac–dc opf re-
garding uncertainty of wind, pv and pev systems. Soft Computing,

𝐶𝑜𝑟𝑟𝑒𝑠𝑝𝑜𝑛𝑑𝑖𝑛𝑔 𝑎𝑢𝑡ℎ𝑜𝑟 𝑚𝑎𝑖𝑙 𝑖𝑑: [email protected] Page 17 of 17