Tariq Kamal12, J.

Electrical Systems x-x (xxxx): xxx-xxx

Uğur Arifoğlu , 1*

Regular paper
Syed Zulqadar
Hassan2
Buck-Boost Converter Small Signal
Model: Dynamic Analysis under System
Uncertainties

An accurate mathematical model of DC-DC converters is an imperative for high performance in all domains of
electronic systems operations. In this work, a small signal circuit model for DC-DC buck-boost converter operated
in continuous conduction mode (CCM) is developed. The proposed modeling is initialized from dynamic equations
illustrating the converter. The non-linear behaviour of the pulse-width modulation and switching process are
addressed via the application of waveform averaging and small-signal modeling techniques. A complete
MATLAB/Simulink model is designed to check the robustness of the proposed converter under different input
voltage and switched load variations. Simulation results present the superiority of proposed model in terms of
transient and steady-state performance, such as small overshoot and short settling time. Furthermore, the
proposed model can be useful to achieve input and output impedances, inductor current variations, and converter
transfer functions to develop a robust closed-loop controller design that can meet stability and performance
conditions of the DC-DC buck-boost converter.

Keywords: Buck-boost converter; state-space average method; continuous conduction mode;

pulse-width modulated converters

1. Introduction
Today's electronic systems need portable, high-quality, modular, reliable, and fast
switching power supplies. Therefore, the widespread applications, (i.e., industrial,
automotive and transportation, electrical machines, renewable energy systems, switching
power supplies and communications, etc) of DC-DC converters have overwhelmed the power
industry. During the modelling of switch mode power supply, ensuring stability is essential.
Normally, a feedback control loop is used for this purpose to obtain the required performance.
However, during the actual operation, various load disturbances and variation of the circuit
component effect the performance of the system. These variations are critical to the behaviour
of switch mode power supply that may result instability. Therefore, in modelling of power
converters, design of controller for these converters is a most thoughtful issue [1], [2].
In literature, various researchers presented different control approaches to regulate the
DC-DC converters, particularly bi-directional DC-DC converters, to acquire a robust output
voltage. The bi-directional DC-DC converters are classified into two: digital simulation
method and analytical modelling method [3]. Those modelling techniques are discussed in
[4]. State-space averaging technique is one of the analytic modelling methods, and relatively
simple with widely applications [5]. Since, DC-DC converters are characterized as non-
linear, and time-variant systems, therefore, control of these converters using linear control

*
Corresponding author: Uğur Arifoğlu, E-mail: [email protected]
1Department of Electrical and Electronics Engineering, Faculty of Engineering, Sakarya University
Serdivan/Sakarya 54050, Turkey, E-mail: [email protected]
2Research Group in Electrical Technologies for Sustainable and Renewable Energy (PAIDI-TEP023), Department

of Electrical Engineering, Higher Polytechnic School of Algeciras, University of Cadiz, Spain.

3State Key Laboratory of Power Transmission Equipment and System Security and New Technology, School of

Electrical Engineering, Chongqing University, Chongqing 400044, China; E-mail: [email protected]

Copyright © JES 2010 on-line: journal/esrgroups.org/jes

 J. Electrical Systems 6-4 (2010): 466-479

techniques is an inefficient. To obtain the prototype of a robust controller, a small-signal

linearized simulated model of the DC-DC converter is essential [6]. Time invariant over a
constant duty cycle is significant benefit of such linearized model. Furthermore, there is no
concern to handle switching and/or switching ripple, and only the DC components of the
waveforms are taken during the modelling. This is obtained through the linearization taken
around a specific operating point from the state space average mode [7], [8].
Moreover, until now various modelling approaches under different software simulation,
viz., PSCAD/EMTDC software [9], internet-based platform PowerEsim [10], PSpice
simulator [11], MATLAB/Simulink [12]–[14], Modelica [15], Scicos simulator [16],
MULTISIM [17], XILINX FPGA [18], LabVIEW/MATLAB [19], of model for DC-DC
converter applications have been appeared in hundreds of papers as well as in many
textbooks to be applied in controller modelling and improve converter’s performance. In [9],
the authors presented a new technique for dynamic modelling of DC-DC buck-boost
converter in continuous conduction mode. The proposed research is based on Laplace and Z-
transforms. Simulations are performed in PSCAD/EMTDC software and its validity is then
supported by experimental. In [10], the researchers applied PSpice simulator and Internet-
based platform PowerEsim for the modelling of switched mode power supply. The design of
Pulse With Modulated (PWM) DC-DC buck-boost converter in the PSpice simulator to
reduce DC error and gain stability is the purpose of [11]. In [12]–[14], the researchers
designed and simulated small signal averaged DC-DC converters in MATLAB/Simulink.
Similarly, study on DC-DC converters using Modelica is developed in [15]. Modelling of
DC-DC converters in Scilab/ScicosLab is attempted in [16]. In [17], the authors designed
DC-DC converter in MULTISIM. The model is tested experimentally with a data acquisition
performed in LabVIEW. Some researchers also presented FPGA based modelling of DC-DC
converter such as in [18]. The authors stated that with the use of PWM and XILINX FPGA
minimizes the cost. In [19], modelling and experimental investigation of closed loop time
domain based discrete PWM DC-DC converter is presented in LabVIEW/MATLAB.
This work presents an AC small-signal linear time-invariant circuit model of a DC-DC
buck-boost converter in operated in CCM. The method is simple and convenient for any
multi-variable system with state-space equations, linear or nonlinear, ideal or non-ideal. İn
proposed method, an approximation is performed in such manner, that the current through
the inductor can return to a steady state value with small overshoot and settling time, and
hence, the true rms value of the currents through the converter components is taken to obtain
the averaged values of waveform. Number of poles and zeros of the system are easily
achievable considering the analytical expressions of the converter transfer functions. The
proposed model is propitious to achieve the expressions for converter transfer function, the
input impedance, the output impedance, and control-to-output voltage transfer function.
This paper is structured as: Section 2 provides methods and materials such as, state-space
average methodology, power stage modelling steps, description about perturbation and
linearization, and canonical form model. Section 3 illustrates proposed small signal Simulink
model. Section 4 describes the simulation results, and conclusion is given in Section 5.
2. Methods and materials
The DC-DC converters are explained into two main types: (1) resonant and soft-switching
converters and (2) hard-switching pulse PWM converters [2]. The PWM converters are
preferred due to its high efficiency, relatively simple control, constant frequency operation,
low component count, and ability to obtain high conversion ratios for both step-down and
step-up applications. The circuit diagram of conventional DC-DC buck-boost converter is
shown in Fig. 1, where D is the main switch, L is input inductor, D is the output diode, and

467
 U. Arifoğlu et al: Buck-Boost Converter Small Signal Model: Dynamic Analysis under sys ...

R is load resistor. Fig. 2 provides the workflow involved in the small signal model. First, the
state space is developed, followed by basic state space averaged modelling. Then, state space
equations are perturbed with small AC variations 𝑑̂ and 𝑣̂𝑖 in the duty cycle and input vector,
and the vectors 𝑥,
̂ 𝑦̂ are the resulting small variations in the output vectors. Next, final state
space averaged model steady state (DC) model and dynamic (AC small signal) model
(linearization) are obtained. Afterwards, the canonical circuit model is established.

Figure 1. Structure of the buck-boost circuit topology.

Figure 2. Flowchart of the small signal modelling approaches

2.1. Modelling of DC-DC Buck-Boost Converter
The state-space averaging approach utilizes the state-space formulation of dynamical
systems to derive the small-signal averaged equations of PWM switching converters [20]. In
this method, first the state space dynamics description of each time-invariant system is
developed. Next, these descriptions are then averaged with respect to their switching period
duration by eliminating the time variance. The final averaged model is nonlinear and time-
invariant. Through linearization process, the model is then linearized around the quiescent
operating point to develop a small-signal model. Then, the model is transformed into
frequency-domain, or s-domain small-signal model, which generates transfer functions. The
derived transfer functions possess all the standard s-domain analysis characteristics and
frequency-domain small-signal dynamics [12]. In state-space averaging technique, a true
state-space equation of the power stage is initially defined. The resulting state-space
description is known as the switched state-space model.
a. Switch ON state period (switched state space model)
The buck-boost converter is obtained from the cascaded integration of step down and the
step-up converters as illustrated in Fig. 1. After closing the switch Q, the diode is reverse

468
 J. Electrical Systems 6-4 (2010): 466-479

biased, and the energy is stored in the coil which is shown in Fig.3. The power stage dynamics
during ON-time period can be represented in the form of a state space equation as:

𝑑𝑥(𝑡)
𝐾 = 𝐴𝑜𝑛 𝑥(𝑡) + 𝐵𝑜𝑛 𝑢(𝑡) (1)
𝑑𝑡

𝑦(𝑡) = 𝐶𝑜𝑛 𝑥(𝑡) + 𝐸𝑜𝑛 𝑢(𝑡) (2)

Eqs. (1-2) can be written in terms of coil voltage, capacitor and source current as:

𝐿 0 𝑑 𝑖(𝑡) 0 0 𝑖(𝑡) 1
[ ] [ ]=[ ][ ] + [ ] [𝑣𝑖 (𝑡)] (3)
0 𝐶 𝑑𝑡 𝑣(𝑡) 0 −1/𝑅 𝑣(𝑡) 0
𝑖(𝑡)
[𝑖𝑖 (𝑡)] = [1 0] [ ] + [0][𝑣𝑖 (𝑡)] (4)
𝑣(𝑡)
b. Switch OFF state period
When the switch is open, the energy stored in the coil is transferred to the load. No energy
is supplied by the source during this interval as shown in Fig. 4. The power stage dynamics
during OFF-time period can be represented in terms of state space equation as:
𝑑𝑥(𝑡)
𝐾 = 𝐴𝑜𝑓𝑓 𝑥(𝑡) + 𝐵𝑜𝑓𝑓 𝑢(𝑡) (5)
𝑑𝑡

𝑦(𝑡) = 𝐶𝑜𝑓𝑓 𝑥(𝑡) + 𝐸𝑜𝑓𝑓 𝑢(𝑡) (6)

Eqs. (5-6) can be written in coil voltage, capacitor current and the source current as:
−𝐿 0 𝑑 𝑖(𝑡) 0 1 𝑖(𝑡) 0
[ ] [ ]=[ ][ ] + [ ] [𝑣𝑖 (𝑡)] (7)
0 𝐶 𝑑𝑡 𝑣(𝑡) 1 −1/𝑅 𝑣(𝑡) 0

𝑖(𝑡)
[𝑖𝑖 (𝑡)] = [0 0] [ ] + [0][𝑣𝑖 (𝑡)] (8)
𝑣(𝑡)

Figure.3. Buck-boost converter equivalent circuit in ON state

Figure.4. Buck-boost converter equivalent circuit in OFF state

469
 U. Arifoğlu et al: Buck-Boost Converter Small Signal Model: Dynamic Analysis under sys ...

2.2. Steady State Solution

The steady state solution is derived by synchronizing the rate of change of dynamic
parameters to zero to obtain the state space averaged balance equations.

0 = 𝐴𝑎𝑣 𝑋 + 𝐵𝑎𝑣 𝑈 (9)

𝑌 = 𝐶𝑎𝑣 𝑋 + 𝐷𝑎𝑣 𝑈 (10)
In Eqs. (9-10), 𝐴𝑎𝑣 , 𝐵𝑎𝑣 , 𝐶𝑎𝑣 and 𝐷𝑎𝑣 are the averaged matrices. Using Eq. 𝐷 + 𝐷𝑜𝑓𝑓 = 1,
𝐴𝑎𝑣 is given as:
0 0 0 1 0 𝐷𝑜𝑓𝑓
𝐴𝑎𝑣 = 𝐷𝐴𝑜𝑛 + 𝐷𝑜𝑓𝑓 𝐴𝑜𝑓𝑓 = 𝐷 [ ] + 𝐷𝑜𝑓𝑓 [ ]=[ ] (11)
0 −1/𝑅 1 −1/𝑅 𝐷𝑜𝑓𝑓 −1/𝑅
Similarly, the averaged matrices 𝐵𝑎𝑣 ,𝐶𝑎𝑣 , and 𝐸𝑎𝑣 are also evaluated as:
1 0 𝐷
𝐵𝑎𝑣 = 𝐷𝐵𝑜𝑛 + 𝐷𝑜𝑓𝑓 𝐵𝑜𝑓𝑓 = 𝐷 [ ] + 𝐷𝑜𝑓𝑓 [ ] = [ ] (12)
0 0 0
𝐶𝑎𝑣 = 𝐷𝐶𝑜𝑛 + 𝐷𝑜𝑓𝑓 𝐶𝑜𝑓𝑓 = 𝐷[1 0] + 𝐷𝑜𝑓𝑓 [0 0] = [𝐷 0] (13)
𝐸𝑎𝑣 = 𝐷𝐸𝑜𝑛 + 𝐷𝑜𝑓𝑓 𝐸𝑜𝑓𝑓 = 𝐷[0] + 𝐷𝑜𝑓𝑓 [0] = [0] (14)
The equilibrium values of the averaged vectors can be derived from Eq. (9) and Eq. (10)
as follows:
−1 𝑖 𝐷𝑉 𝑉
0 𝐷𝑜𝑓𝑓 −
−1 𝐷 [𝑉 ] 𝐼 (𝐷𝑜𝑓𝑓 )2 𝑅 𝑅(1−𝐷)
𝑋 = −𝐴𝑎𝑣 𝐵𝑎𝑣 𝑈 = [ 1 ] [ ] 𝑖 =[ ]=[ 𝐷𝑉 ] = [ 𝐷𝑉 ] (15)
𝐷𝑜𝑓𝑓 − 0 𝑉 − 𝑖 𝑖
𝑅 𝐷𝑜𝑓𝑓 1−𝐷

−1
0 𝐷𝑜𝑓𝑓
𝐷
𝑌 = −(𝐶𝐴𝑎𝑣 −1 𝐵𝑎𝑣 + 𝐸𝑎𝑣 )𝑈 = [𝐼𝑖 ] = − ([𝐷 0] [ 1] [ ] + 0) [𝑉𝑖 ]
𝐷𝑜𝑓𝑓 − 0
𝑅
𝐷 2 𝑉𝑖 𝐷𝑉
=[ ]=[ ] (16)
(𝐷𝑜𝑓𝑓 )2 𝑅 (1−𝐷)𝑅

After any transients have been finished, the input voltage 𝑣𝑖 (𝑡), the input current 𝑖𝑖 (𝑡), the
coil current 𝑖(𝑡) and the capacitor (or load) voltage 𝑣(𝑡) will approach the quiescent points
𝑉𝑖 , 𝐼𝑖 𝐼 and 𝑉, respectively.
where
(1−𝐷)𝑉
𝑉𝑖 = (17)
𝐷
𝐷𝑉
𝐼𝑖 = (18)
𝑅(1−𝐷)
𝑉
𝐼= (19)
𝑅(1−𝐷)
𝐷𝑉𝑖
𝑉= (20)
1−𝐷

2.3. Perturbation and linearization of buck-boost converter

Another way of analyzing easily a non-linear system is its linearization around quiescent
operating point by constructing a small signal model. Fig. 5 illustrates linearization of the

470
 J. Electrical Systems 6-4 (2010): 466-479

V–D curve. The steady state load voltage is V and the duty cycle of the buck-boost converter
is D. Assuming that converter runs in load voltage, V=Vi, coinciding to a quiescent duty
cycle of D = 0.5. The ripples of D (means 𝑑̂ ) around the quiescent point incites the ripples 𝑣̂
in the load voltage V. If the amplitude of 𝑑̂ is enough small, then it can be computed from
the load voltage ripples via linearizing the curve. If the slope of the curve given in Fig.5 is
made equal to the slope of the real curve at the quiescent point value, then the gain of the
DC-DC converter will be equal to the slope value. When ripple of 𝑑̂ is chosen sufficiently
small, the linearized and nonlinear curve will be approximately equal in value.

Figure 5. Linearization of the V(D) curve of the buck-boost converter around quiescent
point for D=0.5.
The ripples in the coil current and load voltage waveforms are eliminated by taking the
average during a period. The components with low frequency of the coil and load waveforms
are written by equations of the form [26].
𝑑<𝑖𝐿 (𝑡)>𝑇
𝐿 =< 𝑣𝐿 (𝑡) > 𝑇 (21)
𝑑𝑡
𝑑<𝑣(𝑡)>𝑇
𝐶 =< 𝑖𝐶 (𝑡) > 𝑇 (22)
𝑑𝑡

where < 𝑥(𝑡) > 𝑇 shows the mean value of x(t) during a period 𝑇. The non-linear averaged
equations of a DC-DC converter can be linearized around a quiescent point. Independent
inputs of the DC-DC converter are taken as constant DC values and AC ripples with low
amplitude. Using Eqs. (21) and (22), the converter nonlinear equations include DC terms,
linear AC terms, and nonlinear terms. If the AC ripples are enough small in amplitude, the
nonlinear variables are much smaller than the linear AC variables, because, they are ignored.
The rest of the linear AC term represents the Small Signal Model (SSM) of the converter.
By averaging the coil voltages and capacitor currents, the basic averaged model which
describes the converter dynamics are given as:
𝑑𝑥(𝑡)
𝐾< > 𝑇 = [𝑑(𝑡)𝐴𝑜𝑛 + 𝑑𝑜𝑓𝑓 (𝑡)𝐴𝑜𝑓𝑓 ] < 𝑥(𝑡) > 𝑇 + [𝑑(𝑡)𝐵𝑜𝑛 + 𝑑𝑜𝑓𝑓 (𝑡)𝐵𝑜𝑓𝑓 ] <
𝑑𝑡
𝑢(𝑡) > 𝑇 (23)
< 𝑦(𝑡) > 𝑇 = [𝑑(𝑡)𝐶𝑜𝑛 + 𝑑𝑜𝑓𝑓 (𝑡)𝐶𝑜𝑓𝑓 ] < 𝑥(𝑡) > 𝑇 + [𝑑(𝑡)𝐸𝑜𝑛 + 𝑑𝑜𝑓𝑓 (𝑡)𝐸𝑜𝑓𝑓 ] <
𝑢(𝑡) > 𝑇 (24)
To design a SSM at a quiescent point (𝐼, 𝑉), it is assumed that the source voltage 𝑣i(𝑡) and
the duty cycle 𝑑(𝑡) are equal to quiescent values 𝑉i and 𝐷, and sum of superimposed AC
variations with small amplitude 𝑣̂i(𝑡) and 𝑑̂ (𝑡). The capital letters show DC components and

471
 U. Arifoğlu et al: Buck-Boost Converter Small Signal Model: Dynamic Analysis under sys ...

the letter with superscript 𝑥̂ are AC components. Perturbation and linearization about a
quiescent point is applied to design the SSM AC model as follows:
< 𝑥(𝑡) > 𝑇 = 𝑋 + 𝑥̂(𝑡)
< 𝑢(𝑡) > 𝑇 = 𝑈 + 𝑢̂(𝑡) (25)
< 𝑦(𝑡) > 𝑇 = 𝑌 + 𝑦̂(𝑡)
𝑑(𝑡) = 𝐷 + 𝑑̂ (𝑡)
where, 𝑢̂, 𝑥̂, 𝑦̂ and 𝑑̂ are small AC variations in control, state, output vectors and duty
ratio, respectively. It is known that AC ripples are much smaller than the quiescent values.
As a result, the state space equations for large signals are obtained as:
𝑑 < 𝑋(𝑡) > 𝑇 𝑑 < 𝑥̂(𝑡) > 𝑇
𝐾( + )
𝑑𝑡 𝑑𝑡
𝑑𝑐 𝑡𝑒𝑟𝑚𝑠
⏞ + 𝐵𝑈 )
= (𝐴𝑋
1𝑠𝑡 𝑜𝑟𝑑𝑒𝑟 𝑎𝑐 𝑡𝑒𝑟𝑚𝑠 (𝑙𝑖𝑛𝑒𝑎𝑟)
⏞
+ (𝐴𝑥̂(𝑡) + 𝐵𝑢̂(𝑡) + ((𝐴𝑜𝑛 − 𝐴𝑜𝑓𝑓 )𝑋 + (𝐵𝑜𝑛 − 𝐵𝑜𝑓𝑓 )𝑈) 𝑑̂ (𝑡)) +
2𝑛𝑑 𝑜𝑟𝑑𝑒𝑟 𝑎𝑐 𝑡𝑒𝑟𝑚𝑠 (𝑛𝑜𝑛𝑙𝑖𝑛𝑒𝑎𝑟)
⏞ − 𝐴 )𝑑̂ (𝑡)𝑥̂(𝑡) + (𝐵 − 𝐵 )𝑑̂ (𝑡)𝑢̂(𝑡))
((𝐴 (26)
𝑜𝑛 𝑜𝑓𝑓 𝑜𝑛 𝑜𝑓𝑓

1𝑠𝑡 𝑜𝑟𝑑𝑒𝑟 𝑎𝑐 𝑡𝑒𝑟𝑚𝑠 (𝑙𝑖𝑛𝑒𝑎𝑟)

𝑑𝑐 𝑡𝑒𝑟𝑚𝑠
⏞
⏞ + 𝐸𝑈) + (𝐶𝑥
𝑌 + 𝑦̂(𝑡) = (𝐶𝑋 ̂(𝑡) + 𝐸𝑢̂(𝑡) + ((𝐶𝑜𝑛 − 𝐶𝑜𝑓𝑓 )𝑋 + (𝐸𝑜𝑛 − 𝐸𝑜𝑓𝑓 )𝑈) 𝑑̂ (𝑡)) +
2𝑛𝑑 𝑜𝑟𝑑𝑒𝑟 𝑎𝑐 𝑡𝑒𝑟𝑚𝑠 (𝑛𝑜𝑛𝑙𝑖𝑛𝑒𝑎𝑟)
⏞ − 𝐶 )𝑑̂ (𝑡)𝑥̂(𝑡) + (𝐸 − 𝐸 )𝑑̂ (𝑡)𝑢̂(𝑡))
((𝐶 (27)
𝑜𝑛 𝑜𝑓𝑓 𝑜𝑛 𝑜𝑓𝑓

The right hands of equations (26) and (27) contain DC (steady state), linear and nonlinear
terms. To obtain a small signal AC model, it is assumed that the DC terms are constant. The
second order AC (nonlinear) terms are much smaller than the first order (linear). If it is
synchronized to each other on both DC sides of the equations (26) and (27), the small signal
linearized state space equations can be written as:
𝑑𝑥̂(𝑡)
𝐾 = 𝐴𝑥̂(𝑡) + 𝐵𝑢̂(𝑡) + (𝐴𝑜𝑛 − 𝐴𝑜𝑓𝑓 )𝑋 + ((𝐴𝑜𝑛 − 𝐴𝑜𝑓𝑓 )𝑋 + (𝐵𝑜𝑛 − 𝐵𝑜𝑓𝑓 )𝑈) 𝑑̂ (𝑡)
𝑑𝑡
(28)

𝑦̂(𝑡) = 𝐶𝑥̂(𝑡) + 𝐸𝑢̂(𝑡) + ((𝐶𝑜𝑛 − 𝐶𝑜𝑓𝑓 )𝑋 + (𝐸𝑜𝑛 − 𝐸𝑜𝑓𝑓 )𝑈) 𝑑̂ (𝑡) (29)

The resultant small signal AC equations of the ideal buck-boost converter are given as:
0 𝐷𝑜𝑓𝑓
−𝐿 0 𝑑 𝑖̂(𝑡) 𝑖̂(𝑡) 𝐷
[ ] [ ]=[ 1 ][ ] + [ ] 𝑣̂(𝑡)
0 𝐶 𝑑𝑡 𝑣̂(𝑡) 𝐷𝑜𝑓𝑓 − 𝑣̂(𝑡) 0 𝑖
𝑅
0 −1 𝐼 1
+ ([ ] [ ] + ([ ]) 𝑉𝑖 ) 𝑑̂ (𝑡) (30)
−1 0 𝑉 0
𝑑𝑖̂ (𝑡)
−𝐿 = 𝐷𝑜𝑓𝑓 𝑣̂(𝑡) + 𝐷𝑣̂(𝑡)
𝑖 + (𝑉𝑖 − 𝑉)𝑑̂ (𝑡) (31)
𝑑𝑡
𝑑𝑣̂(𝑡) 1
−𝐶 = 𝐷𝑜𝑓𝑓 𝑖̂(𝑡) − 𝑣̂(𝑡) − 𝐼𝑑̂ (𝑡) (32)
𝑑𝑡 𝑅

472
 J. Electrical Systems 6-4 (2010): 466-479

𝑖̂(𝑡)
𝑖̂(𝑡)
𝑖 = [𝐷 0] [ ] + [0]𝑣̂(𝑡)
𝑖 + (([1 0] − [0 0])𝑋 + ([0] − [0])𝑉𝑖 )𝑑̂ (𝑡) (33)
𝑣̂(𝑡)
𝑖̂(𝑡) 𝐼
𝑖̂(𝑡)
𝑖 = [𝐷 0] [ ] + (([1 0]) [ ]) 𝑑̂ (𝑡) (34)
𝑣̂(𝑡) 𝑉
Fig.6 shows the resultant small signal model of the ideal buck-boost converter.

Figure 6. SSM of the buck-boost converter into canonical form

2.4. The Canonical Circuit Model
The alternating current equivalent circuit of any continuous current mode DC-DC
converter can be expressed in the canonical form [15].

Figure 7. Obtaining the SSM of the buck-boost converter into canonical form step by step

473
 U. Arifoğlu et al: Buck-Boost Converter Small Signal Model: Dynamic Analysis under sys ...

The circuit model in Fig.6 is achieved using equations (31-34). Fig.7, explains developing
of the canonical model of essential buck boost DC-DC converter in four steps. The turn ratio
of the transformer shown in step 4 in Fig. 7 is 𝐷𝑜𝑓𝑓 ∶ 𝐷 or 1: D/(1 − 𝐷). This conversion
ratio is a function of the quiescent duty cycle 𝐷. In addition, the control function of the model
contains dependent voltage P(𝑠)∗ 𝑑̂ (𝑠) and dependent current source J(𝑠)*𝑑̂ (𝑠) in Fig.7. Both
sources are driven by 𝑑̂ . Model developed in Step 4 in Fig.7 is an equivalent circuit that shows
the low frequency small signal variations in the converter waveforms. It can now be solved,
using conventional linear circuit analysis techniques, to find the converter transfer functions,
output impedance, and other AC quantities of interest.
3. Description of Proposed Simulink Model
The proposed Simulink model is shown in Fig. 8. The circuit consists of two controlled
voltage sources, a controlled current source, an ideal transformer, a low pass filter and the
time varying resistive load. The controlled voltage source 1 (CVS1) is DC input voltage of
the DC-DC buck-boost converter, it generates variable step voltages without ripples (case 1)
when connected to the signal builder block named variable voltage without ripple. Similarly,
it generates variable step voltages with ripples (case 2), when connected to the embedded
MATLAB block named variable voltage with ripples to end "s" of CVS1 in Fig.7. The
controlled current source and CVS2 represent the control function of the SSM of DC-DC
buck-boost converter. The dependent sources are related to duty cycle. The ideal transformer,
an imaginary device, is widely used in DC-DC power conversion circuits to change the levels
of voltage and current waveforms while transferring electrical energy. This block can be used
to represent either an AC transformer or a solid-state DC-DC converter.

Figure 8. Proposed Simulink model of the buck-boost DC-DC converter

4. Simulation studies
The performance of the proposed DC-DC buck-boost converter model is checked at
different input voltage and load variations in MATLAB/Simulink.
Case 1: Step changes in input voltage without ripple and the time varying load
Fig.9 shows variations at different intervals in input voltage and load without ripple. The
simulated response of the load current due to step change in input voltage and the load is
given in Fig.10. The simulated response of the load voltage due to step change in the input
voltage and the load is shown in Fig.11. The simulated response of the inductor current due
to step change in the input voltage and the load is given in Fig.12. The simulated response of

474
 J. Electrical Systems 6-4 (2010): 466-479

the capacitor current due to step change in the input voltage and the load is also shown in
Fig. 13. From these figures, output voltage tracks the input voltage with the prescribed
quiescent duty cycle and the output voltage and inductor current can return to steady state
value with small overshoot and settling time. The capacitor current almost zero value except
in the instant of input voltage and load change.

Figure 9. Input voltage variation and load profile form 4 Ω to 8 Ω

Figure10. Simulated response of the load current due to step change in input voltage and load

Figure11. Simulated response of the load voltage due to step change in the input voltage
and load

475
 U. Arifoğlu et al: Buck-Boost Converter Small Signal Model: Dynamic Analysis under sys ...

Figure12. Simulated response of the inductor current due to step change in the input voltage
and load

Figure13. Simulated response of the capacitor current due to step change in the input voltage
and load

Figure14. Input voltage variation (with ripple)

Case 2: Step changes in input voltage with ripple and the time varying load
Fig.14 illustrates the input voltage variations with ripple having 3 steps from 24 V to 5 V
at 10 sec, and at the same time the load variations from 4 Ω to 8 Ω at 10 sec are also applied
to the proposed model. The simulated response of the load voltage due to step change in input
voltage and the load is given in Fig.15. The simulated response of the load current due to step
change in the input voltage and the load is given in Fig.16. The simulated response of the
inductor current due to step change in the input voltage and the load is provided in Fig.17.
From these figures, output voltage tracks the input voltage with the prescribed quiescent duty
cycle and the output voltage and inductor current can return to a steady state value with small

476
 J. Electrical Systems 6-4 (2010): 466-479

or large overshoot and settling time. For both of the cases, D=0.5, Le= (30e-6) *1/((1-D)^2)
Henry, C=2.2 e-3 Farad, d=0.001, Ts=1e-5 are selected in Fig. 8.

Figure.15. Simulated response of the load voltage due to step change in the input voltage
(with ripple) and the load

Figure.16. Simulated response of the load current due to step change in input voltage (with
ripple) and the load

Figure.17. Simulated response of the inductor current due to step change in the input
voltage (with ripple) and the load
5. Conclusion
This work not only provides the benefits of state-space averaging modelling, but also give
the necessary steps and ways of building a dynamic model of the system for an application
or algorithm analysis. The proposed model was checked in CCM under load changes and
input voltage variations. Results were shown to indicate that the output voltage and inductor
current can return to steady state with a small overshoot and settling time even under
disturbances. The proposed model was shown to be very accurate in anticipating the large or
low signal time domain transients. In addition, this model can also provide a new control
technique to study closed loop performance of a system. The model can be used to develop

477
 U. Arifoğlu et al: Buck-Boost Converter Small Signal Model: Dynamic Analysis under sys ...

strong, and robust closed loop controller, which can ensure stability and performance
conditions of the DC-DC buck-boost converter, and can be beneficial to obtain circuit
parameters, such as input and output impedances, inductor current variations, the converter
transfer functions, etc.

Acknowledgment
The authors wish to thank Turkish Government Scholarship 2015-2016 for financial
support during the completion of this research work.

References

[1] M. Sarvi, I. Soltani, N. Namazypour, and N. Rabbani, “A new sliding mode controller for DC/DC
converters in photovoltaic systems,” J. Energy, vol. 2013, 2013.
[2] H. R. Muhammad, “Power electronics handbook,” Acad. Pr, 2001.
[3] W. Zhang, X. Tan, K. Wang, T. Li, and J. Chen, “Design for boost DC-DC converter controller based on
state-space average method,” Integr. Ferroelectr., vol. 172, no. 1, pp. 152–159, 2016.
[4] X. Lei and C. GUILAN, “Stability analysis of BOOST converter based on the state space average model
[J],” Sci. Technol. Inf., vol. 3, pp. 257–258, 2009.
[5] I.-Y. Chung, W. Liu, K. Schoder, and D. A. Cartes, “Integration of a bi-directional DC–DC converter
model into a real-time system simulation of a shipboard medium voltage DC system,” Electr. Power Syst.
Res., vol. 81, no. 4, pp. 1051–1059, 2011.
[6] M. R. Modabbernia, F. K. Khoshkbijari, R. Fouladi, and S. S. Nejati, “The state space average model of
buck-boost switching regulator including all of the system uncertainties,” Int. J. Comput. Sci. Eng., vol.
5, no. 2, p. 120, 2013.
[7] A. Reatti and M. K. Kazimierczuk, “Small-signal model of PWM converters for discontinuous conduction
mode and its application for boost converter,” IEEE Trans. Circuits Syst. I Fundam. Theory Appl., vol.
50, no. 1, pp. 65–73, 2003.
[8] J.-H. Su, J.-J. Chen, and D.-S. Wu, “Learning feedback controller design of switching converters via
MATLAB/SIMULINK,” IEEE Trans. Educ., vol. 45, no. 4, pp. 307–315, 2002.
[9] H. M. Mahery and E. Babaei, “Mathematical modeling of buck–boost dc–dc converter and investigation
of converter elements on transient and steady state responses,” Int. J. Electr. Power Energy Syst., vol. 44,
no. 1, pp. 949–963, 2013.
[10] G. Marinova, “PSpice as a verification tool for switch mode power supply design with PowerEsim,” in
Nonlinear Dynamics of Electronic Systems, Proceedings of NDES 2012, 2012, pp. 1–4.
[11] J. J. Lee, “Analysis of a Small-Signal Model of a PWM DC-DC Buck-Boost Converter in CCM.” Wright
State University, 2007.
[12] A. A. Elbaset and M. S. Hassan, “Small-Signal MATLAB/Simulink Model of DC–DC Buck Converter,”
in Design and Power Quality Improvement of Photovoltaic Power System, Springer, 2017, pp. 97–114.
[13] X. L. Xiao, X. J. Lu, J. W. Yi, and X. H. Ding, “Small-Signal Equivalent Circuit Modeling and Controller
Design of DC/DC Converter Based on Matlab,” in Advanced Materials Research, 2015, vol. 1070, pp.
1586–1591.
[14] S. K. Kollimalla and M. K. Mishra, “Analysis and control of buck-boost converter using average power
balance control (APBC),” in India Conference (INDICON), 2012 Annual IEEE, 2012, pp. 161–166.
[15] D. A. Torrey and U. S. Selamogullari, “A Behavioral Model for DC-DC Converters using Modelica,” in
Proceedings 2nd International Modelica Conference, 2002.
[16] B. P. Meneses, M. T. S. Quixadã, J. C. T. Campos, and L. L. N. dos Reis, “Simulation of converters using
Scilab/Scicoslab,” in Power Electronics Conference (COBEP), 2011 Brazilian, 2011, pp. 1032–1036.
[17] J. I. CORCAU and L. DINCA, “NUMERICAL MODEL AND EXPERIMENTAL TESTS FOR DC TO
DC BOOST CONVERTER.,” DAAAM Int. Sci. B., 2015.
[18] P. Karthigeyan, M. S. Raja, T. S. Kumar, S. R. S. Ganesh, and J. Lavanya, “Simulation of Bi-directional
DC-DC Converter Using FPGA,” Procedia Comput. Sci., vol. 79, pp. 708–714, 2016.
[19] S. R. Raja, “Time Domain Based Digital Controller for Buck-Boost Converter,” J. Electr. Eng. Technol.,
vol. 9, no. 5, pp. 1551–1561, 2014.
[20] R. W. Erickson and D. Maksimovic, Fundamentals of power electronics. Springer Science & Business
Media, 2007.

478