Research Article GU J Sci 34(x): x-x (2021) DOI: 10.35378/gujs.

947882
Gazi University

Journal of Science
http://dergipark.gov.tr/gujs

Robust Reference Tracking and Load Rejection on Non-Linear System Using

Controllers

Renjini G.S1,* , Deepa THANGAVELUSAMY2

1
Industrial Electric & Control Systems, SVG Electric & Control Products, Perumal Kovil Street, Barur (PO), Krishnagiri (Dt), Tamil Nadu,
India
2
SELECT, VIT University, Chennai, India.

Highlights
• This paper focuses on different control topologies of power converter
• A Non-Linear system was investigated in steady state and transient region.
• Proposed system was regulated the DC output voltage in load disturbances region.

Article Info Abstract

DC-DC converters form a class of highly nonlinear systems and are widely used in most practical
Received: 4 June 2021 applications. Positive Output Elementary Luo Converter (POELC) is a fourth-order converter
Accepted: 17 Oct 2021 used to regulate dc voltages in low power applications with varying input voltage and load
conditions. POELC is employed to regulate dc voltages for sensitive load applications like SMPS,
powering hard disks etc. The control of higher-order converters is always a challenging task due
Keywords to the highly nonlinear nature of these converters. Hence the development of an optimal controller
Positive output elementary for POELC is proposed in this paper. The proposed control methods are investigated for transient
luo converter, and steady-state performances considering practical operating conditions. The controllers are
Proportional integral designed, simulated and the performance of POELC with various controllers is evaluated in the
controller, MATLAB/Simulink environment. The PI controller exhibits better performance in the linear
Linear quadratic region whereas, in the nonlinear region, due to time variance and switching nature. To minimize
regulator, the sensitivity to external disturbances and to improve the stability margins, an LQR is designed
Integral absolute error,
to regulate the output of POELC. The result shows that a well-designed LQR feedback can
Integral square error
stabilize the system and enhances its performance in terms of Integral Absolute Error (IAE),
Integral Square Error (ISE). Certain case studies are discussed concerning practical applications.

1. INTRODUCTION

DC-DC converters are widely used in battery power applications in many areas. For multiple battery
mission requires large storage space and also its stored energy starts to decay when the voltage is drained.
This drawback is overcome by using switched DC-DC converters. The second-generation converters utilize
the positive, negative and dual-output of single quadrant Luo converter to acquire the bidirectional power
flow with higher efficiency. In addition with a novel geometric progression was introduced for the lifting
of voltage to achieve high voltage gain using a simple structure in a cascade converter. Besides with
conventional converter filter components are connected to reduce radio frequency interference. The
advantage of this proposed converter has been verified by comparing it with existing converters via
simulation and hardware setup [1, 2].

The Numerous control approaches and tuning methods are used to regulate the load voltage constant for
servo as well as regulatory responses. The conventional PI control techniques are commonly used in much
Industrial application. Combination of proportional and integral gain constitutes PI controller. Zeigler
Nichols (ZN) is one of the most prevalent methods used to determine the proportional gain and the Integral
gain [3-6]. This controller provides stable operation against load disturbances and minimizes the saturation
risk. The unstable equilibrium point is removed from the balance dynamics. PI controller for POESLL

*Corresponding author, e-mail: [email protected]

 Renjini G.S, Deepa THANGAVELUSAMY / GU J Sci, 36(1): x-x (2023)

converter to attain static and dynamic performance over Proportional Integral and Derivative (PID)
controller. The mathematical modelling of the converter and controller are performed by the state-space
averaging method [7]. PI Controller provides poor performance for Non-linear system. So this kind of
system requires some expertise knowledge like fuzzy logic for better performance system. From the
literature survey [8-10], comparative study of FLC based PI controller with another controller in terms of
transient response, steady-state error, disturbance rejection from the load and current variations are studied
in different closed-loop systems. These feedback systems provide regulated boosted output voltage. The
simulation reveals that superior performance is exhibited by the proposed controller. But these controllers
do not forecast the future behaviour of the Non-linear system. Hence predictive voltage controller is used
to suppressing cross-regulation issues in a single-input multiple-output buck converter. The converter
operates in Continuous Conduction Mode (CCM) and has quick dynamic responses, adaptable to variable
switching frequency [11]. This controller can improve the steady-state error at low frequency. Even though
with simple construction, less computation and fixed frequency, this controller is not robust against load
variation and signal disturbances and also this system provides poor performance indices for load
disturbance regions [12].

Most of the early works done in this area focused more on finding PID controllers that stabilize the nominal
plant model. Many authors continued their research by using intelligent control techniques like Fuzzy, SMC
etc. Some of the authors extended their research work to optimize the PID controller parameter which will
meet certain design criteria and provide robustness. The beginning of PID controller is more than five-
decade-old and at present, its functionality has been furthermore improvised with an aid of advanced control
techniques like Model Predictive Controller, Robust controller etc. which deal with issues like stability,
efficient computation, optimization, constraints etc. But the control of the higher-order plant is always a
challenging task due to its highly nonlinear nature. System designed using optimal LQR balances between
acceptable response and the amount of control energy required [13].

Find a feasible control, such that the system that starts from the given initial condition transfer its state to
the objective set, and in so doing minimizes a performance index. In industrial systems, there are some
situations such as the improving optimal use of available resources, chemical process, traffic control
systems and robotic systems where optimal control can be applied, such as the control of bacterial content
in a bioengineering system [14].

Linear Quadratic Regulator shows better performance indices for the system which undergoes various
uncertain parameters like source voltage, load and duty ratio. The weight matrices of the cost function are
determined using various methods. It determines the error and energy expense rate of the system. Compared
with the existing methods, this method shows superior performance like less settling time, minimum
amplitude control signal with less error. The results are not verified experimentally to validate the
simulation performance [15, 16]. The dynamic parameters such as source and reference voltages along with
load current are varied to measure the response of the controllers. The peak overshoot, undershoot and
reverse recovery time are measured to find the competence of the LQR. The simulation models were
developed to analyze the transient and steady-state performances. Literature surveys enlighten the necessity
of the controllers for the Non-linear system [17].

In this paper, Positive Output Elementary Luo Converter (POELC) is considered as a Single Input Single
Output (SISO) system. POELC provides good steady-state performance characteristics. But its dynamic
performance is affected due to the fourth-order characteristics which lead to closed-loop bandwidth
limitation for large-signal stabilization. Hence conventional PIC and LQR techniques are incorporated to
converter for better performance. Section 2 details the power converter operation, descriptions and design
parameters. Section 3 and 4 illustrate the control topologies such as PIC and LQR. Section 5 describes the
simulation results and Section 6 details the conclusion of this paper.

2. POWER CONVERTER

Power Electronics is an emerging technology that provides an interface between the source and the load to
transfer power. Most of the utility mains require very low voltages internally. This low voltage is achieved
 Renjini G.S, Deepa THANGAVELUSAMY / GU J Sci, 36(1): x-x (2023)

by the use of Switched Mode Power Supplies (SMPS). The major role of DC-DC converters is to transfer
power from the source to load. The Switched-mode DC-DC converters are used for converting voltage from
one level to another by switching action. They are widely used in SMPS and DC motor drive applications.
The output voltage of the converter should be regulated against source voltages and load perturbations. By
switching the on and off duration of the switching pulse, the converter generates an average output dc
voltage. But there are some drawbacks to the fundamental DC-DC converter used in SMPS. The efficiency
of the converters is very poor for high gain and also it is very difficult to control and analyze. The control
of higher-order converters is always a challenging task due to the highly nonlinear nature of these
converters. Moreover, the effect of noise and Electro-Magnetic Interference (EMI) is quite significant, thus
EMI filters are also required. Positive Output Elementary Luo Converter (POELC) overcomes these
drawbacks and also it is efficiently used in SMPS for 200W [18].

Figure 1. Positive output elementary Luo converter

Figure 1 shows the circuit diagram of POELC. This is a direct positive output buck-boost converter. The
components consist of inductors L1, L2, capacitors C1, C2, freewheeling diode D, and the power switch S.
The load resistance RL is working on the switching frequency Fs with duty ratio δ for source voltage Vin and
the output voltage Vo. It consists of, S − L1 − C1 − D act as pump circuit and L2 - C2 act as low pass filter.
Capacitor C1 has a sufficiently large value to store and transfer energy. The converter operates in two modes
[19].

2.1. Switch on Condition

Figure 2 illustrates POELC in on mode. When the switch is closed, the current flowing through the inductor
L1 starts increasing for the supply voltage. Energy stored in capacitor C1 is transferred to the load through
the inductor L2. The load is supplied by capacitor C2. Since the diode D is reverse biased, the inductor
currents iL1 and iL2 increase until the switch S is turned off. The Equations (2) to (5) are derived by using
Kirchoff Voltage Law (KVL) and Kirchoff Current Law (KCL) for POELC during the switch-on condition.

Figure 2. POELC-Switch on mode

1 ; 0 ≤ 𝑡 ≤ 𝑘𝑇
𝜕={ } (1)
0; 𝑘𝑡 ≤ 𝑡 ≤ 𝑇

𝑑𝑖𝐿1 𝐸
𝑑𝑡
=𝐿 (2)
1

𝑑𝑖𝐿2 𝐸−𝑉𝑐1 −𝑉𝑐2

𝑑𝑡
= 𝐿2
(3)
 Renjini G.S, Deepa THANGAVELUSAMY / GU J Sci, 36(1): x-x (2023)

𝑑𝑉𝑐1 𝑖𝐿2
=− (4)
𝑑𝑡 𝐶1

𝑑𝑉𝑐2 𝑖𝐿2 𝑉𝑐2

𝑑𝑡
= 𝐶2
− 𝑅𝐶2 (5)

where
iL1 - The inductor current L1,
iL2 - The inductor current L2,
Vc1 - The capacitor voltage C1,
Vc2 - The capacitor voltage C2,
Vin - Input voltage.

2.2. Switch off Condition

Figure 3 implies that the energy stored in the inductors L1 and L2 is transferred to the capacitor C1 and C2,
R through the respectively. Freewheeling diode D is used for continuous conduction mode. Equations (6)
to (9) explains the circuit dynamics under mode 2 are obtained by applying the KCL and KVL.

Figure 3. POELC-Switch off mode

𝑑𝑖𝐿1 𝑉𝑐1 (6)

=−
𝑑𝑡 𝐿1

𝑑𝑖𝐿2 −𝑉𝑐2 (7)

𝑑𝑡
= 𝐿2

𝑑𝑉𝑐1 𝑖𝐿1
= (8)
𝑑𝑡 𝐶1

𝑑𝑉𝑐2 𝑖𝐿2 𝑉𝑐2

𝑑𝑡
= 𝐶2
− 𝑅𝐶2
. (9)

The final state space equivalent matrices A, B, C, D of the converter derived from the Equations (2) to (9)
are given below
 Renjini G.S, Deepa THANGAVELUSAMY / GU J Sci, 36(1): x-x (2023)

𝑑𝑉 𝑐2 (𝑡) 1 1 𝑉𝑐2 0 (10)

− 𝑅𝐶 𝐶2
0 0
𝑑𝑡 2
1 𝜕 𝜕
𝑑𝑖𝐿2(𝑡)
−𝐿 0 0 𝑖𝐿2 𝐿2
𝑑𝑡 2 𝐿2
𝑑𝑉𝑐1(𝑡)
= 𝜕 (1−𝜕) + 𝑉𝑖𝑛 .
0 −𝐶 0 𝑉𝑐1 0
𝑑𝑡 1 𝐶1
𝑑𝑖𝐿1(𝑡) (1−𝜕)
[ ] [ 0 0 0 ][𝑖 ] 𝜕
𝑑𝑡 𝐿1 𝐿1 [𝐿1 ]

2.3. Design of Converter Components

The mathematical model of the converter is derived by the state space average method in which the state
equation is derived for the power switch conduction period and the diode conduction period. Assuming the
design parameters of the POELC for output voltage Vo = 24V for the proposal values of input voltage
Vin = 12 V, Frequency Fs= 50 KHz, Inductor L1=L2 =1mH, Capacitors C1= C2= 10µF, Duty cycle 𝜕 = 0.5,
Resistance RL = 2.8Ω, Output Power Po = 200 W.

DC voltage conversion ratio M

V (11)
M = Vo = 1 .
in

The ripple current in inductor L1

(1−𝜕)RL (12)
ε1 = 2MFs L1
= 0.1 .

The ripple current in inductor L2

𝜕RL (13)
ε2 = = 0.4 .
2MFs L2

The ripple voltage in Capacitor C1

𝜕 (14)
ρ1 = 2F = 0.0025 .
s RL C1

The ripple voltage in Capacitor C2

δ (15)
ρ2 = 8MF 2 = 0.00125 .
s C2 L2

Table 1. Design Parameters of POELC-SMPS

Components Inductor Inductor Capacitor Capacitor Resistor Switching Input
L1 L2 C1 C2 RL Frequency Voltage
Fs Vin
Ratings of 1mH 1mH 10µF 10µF 2.8 Ω 50KHz 12V
Components

In POELC the power stage contains the inductors and capacitors which introduce nonlinearity and their
behaviour cannot be judged easily. Therefore there is a necessity to linearize a nonlinear system.
Linearization is a method for assessing the local stability of an equilibrium point of a system of nonlinear
differential equations or discrete dynamical systems [20]. To linearize the system, small perturbations in
 Renjini G.S, Deepa THANGAVELUSAMY / GU J Sci, 36(1): x-x (2023)

the variables are considered. From 0.4 to 0.5 duty cycle, around the operating point, the converter exhibits
linear characteristics. Hence a duty cycle of 0.5 is chosen. The design values for POELC components are
tabulated in Table 1. These values are substituted in the system state space matrix Equations (2) to (9) and
the output to control input transfer function is obtained as,

𝑉𝑜 (𝑠) 4.86∗1010 𝑠2 −0.00017𝑠+7.29∗1020 (16)

= 𝑠4 +5.79∗105𝑠3 +8.4∗1010𝑠2 +8.5∗1015𝑠+3.13∗1020 .
𝑉𝑖𝑛 (𝑠)

Zeros are

s1 = 0.0000 + 122470i, s2 = 0.0000 - 122470i.

Poles are

s1 = -423280, s2 = -48670 + 101700i, s3= -48670 - 101700i, s4 = -58080.

It is observed that the open-loop system has four negative poles out of which two are negative real poles
and the other two are complex poles. Hence the system is stable. Even though it is a stable system, it exhibits
nonlinear behaviour when subjected to disturbances. These parameters change with time causing the
converter operating point to change. Hence the converter operation may deviate from the steady-state
condition. Controllers are incorporating into the power converter for obtaining stabled controlled operation
for a wide operating range and are also used to achieve satisfactory static and dynamic performance [21].
Non-zero error is important for any closed-loop system. Hence to find Integral Absolute Error (IAE) to be
more appropriate. In the case of Integral Square Error (ISE), the square-of-error converges to zero. The
main objective function is to optimize the controller parameters that minimize the IAE and ISE criteria.

Mathematically it is represented by,

∞
𝐼𝐴𝐸 = ∫0 𝑎𝑏𝑠(𝑟(𝑡) − 𝑦(𝑡))𝑑𝑡 (17)

∞
𝐼𝑆𝐸 = ∫ (𝑟(𝑡) − 𝑦(𝑡))2 𝑑𝑡 (18)
0

where r(t) - reference input signal, y(t) - measured output signal.

3. PI CONTROLLER

Proportional Integral Controller (PIC) is a feedback controller which obtains the error and determines the
output based on the characteristics of the error signal. The difference between the reference variable r(t)
and the process variable y(t) is given as error signal e(t) which are evaluated and measured to get the output.
This output signal u(t) is given as the controlled input for the converter. Proportional Controller (PC)
introduces the steady-state error. The increase in gain reduces the error. But it makes the system oscillatory.
The consequence of summing integral controller introduces a pole and reduces the steady-state error [22].
The block diagram of the PIC designed for POELC is represented in Figure 4.

Figure 4. Block diagram of PIC for POELC

 Renjini G.S, Deepa THANGAVELUSAMY / GU J Sci, 36(1): x-x (2023)

The structure of the PI control system is

𝑡
𝑢(𝑡) = 𝐾𝑝 𝑒(𝑡) + 𝐾𝑖 ∫ 𝑒(𝑡)𝑑𝑡 (19)
0

𝐾𝑝 𝑡
𝑢(𝑡) = 𝐾𝑝 𝑒(𝑡) + ∫ 𝑒(𝑡)𝑑𝑡 (20)
𝑇𝑖 0

where e(t) = r(t) − y(t) ; u(t) - controlled input signal,e(t) - error signal.

For the past five decades, several tuning methods were developed for calculating the PIC parameters. One
of the renowned tuning procedures is ZN method. In this method, the proportional gain is varied by keeping
the Integral value as zero-till sustained oscillations are obtained. The step response of the open-loop transfer
function is presented in Figure 5. Appling ZN method, time delay L=0.000004s and the time constant
a=0.28 is observed. Substituting these values into equation Kp = 0.9/a=3.214 ,
Ti= 3L=0.012 ms are calculated

1
𝐾(𝑠) = 𝐾𝑝 (1 + )
𝑇𝑖 𝑠 (21)

267857
𝐾(𝑠) = 3.214 + . (22)
𝑠

Figure 5. Step response of the open loop transfer function

 Renjini G.S, Deepa THANGAVELUSAMY / GU J Sci, 36(1): x-x (2023)

Figure 6. Servo Response of PIC for POELC-SMPS

The optimal setting values of PIC are attained by the ZN method for desired performance. Figure 6 shows
the servo response of PIC for POELC. The reference voltages or Set Points (SP) are set as 24V at 0s, 29V
at 1s and 24V at 2s of interval. The Control Variable (CV) and the Manipulated Variable (MV) for the
corresponding reference voltages are mentioned graphically.

Figure 7. Regulatory Response of PIC for POELC-SMPS

Figure 7 depicts the regulatory response of PIC for POELC. The load is increased and decreased by 20%
of the interval of 1s. Responses show oscillation of higher amplitudes and later on, it subsides with time.
The settling time of the response is more.

4. LINEAR QUADRATIC REGULATOR

Linear Quadratic Regulator (LQR) is the linear controller mainly designed for better quadratic
performances. The system designed by using optimal controller where the closed-loop poles are allocated
such that dominant poles move towards the Left Half of the s-Plane (LHP) then the remaining poles are
non-dominant. Hence it makes the system less sensitive in the disturbance regions.

Figure 8. Block diagram of LQR for POELC

 Renjini G.S, Deepa THANGAVELUSAMY / GU J Sci, 36(1): x-x (2023)

State-space modelling is used for the easy representation of the system. This controller has performed well
in both steady- state and transient conditions [23]. Feedback control law is defined by,

u = −Kx(t). (23)

In this technique, the feedback gain matrix (K) is determined by optimizing J energy function
∞
J = ∫0 (x T Qx + uT Ru)dt . (24)

Matrices Q and R express the relationship between error and energy expense rate. Optimal control gain K
is given by,

K = R−1 B T P . (25)

Thus the optimal control law is given by,

u = −R−1 B T P x(t). (26)

In addition to the state feedback gain K, LQR returns the solution P of the associated reduced matrix Riccati
equation is given as follows,

AT P + PA − PBR−1 B T P + Q = 0 . (27)

Figure 8 shows the block diagram of LQR in POELC. From the Equations (2) to (9), the system matrices
A, B, C, D of the POELC for the corresponding design values mentioned in Table 1 are as follows

A=
−578703 1666666 0 0
−41667 0 29167 0
[ ]
0 −350000 0 150000
0 0 −30000 0

0
29167
B= [ ]
0
70000
C = [1 0 0 0]

D = [0] .

To minimize the performance index J by using the weighting matrices Q and R subject to the constraints,

x = Ax(t) + Bu(t) (28)

x(𝑡0 ) = 𝑥0 (29)

y = Cx(t) . (30)
 Renjini G.S, Deepa THANGAVELUSAMY / GU J Sci, 36(1): x-x (2023)

State variables are represented by,

𝑥1 = 𝑣𝑐2
𝑥2 = 𝑖𝐿2
𝑥3 = 𝑣𝑐1
𝑥4 = 𝑖𝐿1 .

Output matrix C denotes that x1 is accountable for the servo response of the system such that the optimal
gain matrices K(1) is connected through feed-forward for the system. The regulatory response is controlled
by x2, x3, x4. The corresponding gain matrixes K(2), K(3) and K(4) are connected as feedback for the system.
Solution of Riccati Equation (13), P matrix is given by,

0.1133 0.0011 0.0000 0.0000

0.0011 0.0023 0.0000 0.0009
P = 1.0e−5 × [ ].
0.0000 0.0000 0.0003 0.0000
0.0000 0.0009 0.0000 0.0004

Positive-definite matrix P will exist only for a stable system. Therefore matrix A-BK is stable. By
substituting this matrix P into Equation (25), optimal gain matrix K is obtained as detailed below

K = [314150 5990 0 0] .

For a stable system, closed loop characteristic equation becomes, AA=A-BB=0, where BB=B*K

0.0001 0.0002 0 0
0.9163 0.0175 −0.0000 0.0000
AA = 1.0e+10 × [ ]
0 0.0000 0 0.0000
2.1990 0.0419 −0.0000 0.0000
0 0 0 0
0.9163 0.0175 0.0000 0.0000
BB = 1.0e+10 × [ ].
0 0 0 0
2.1990 0.0419 0.0000 0.0000

Numeral methods are used to find the weighting matrices Q and R. The weight matrices of the cost function
are determined using trial and error method. Q and R are the positive-definite matrices which are given by,

100 0 0 0
0 0 0 0
Q=[ ]
0 0 0 0
0 0 0 0

R = [0.000000001];

−8.7671 + 8.7671i
+07 −8.7671 − 8.7671i
E = 1.0e ×[ ].
−0.0000 + 0.0122i
−0.0000 − 0.0122i

The LQR engages the choosing of Q and R matrix to minimize the energy function which offers adequate
performance. The closed-loop eigenvalues were interrelated. This system performance is evaluated in terms
of Root locus approach. Figures 9 and 10 illustrate the pole-zero plot of the uncontrolled and controlled
 Renjini G.S, Deepa THANGAVELUSAMY / GU J Sci, 36(1): x-x (2023)

plant. The open-loop system has four negative poles out of which two are negative real poles and the other
two are complex poles.

Figure 9. Pole-Zero plot for uncontrolled plant

Figure 10. Pole-Zero plot for controlled plant

According to the control theory, if two real poles located near ‘jw’ axis and its magnitude are 5 to 10 times
lesser than the real pole located far away from the axis with high amplitude then the former one will be the
dominant pole. If the system has a dominant complex pole that affects the transients then it will decay
slowly. In pole placement method, poles must be located to the open left half of the complex plane. Hence
the location of the closed loop poles on the complex plain is quiet difficult, and requires high level of skills
in establishing relationship between poles and dynamic performances of the closed loop control system .If
the system equipped with LQR, then it moves the dominant complex poles towards the zeros thereby pole-
zero cancellation takes place. So the system is approximated to the second-order system. Remaining poles
will become non-dominant. Hence the system provides improved dynamics in both steady-state and
transient regions [24]. Figure 11 shows the servo response of LQR for POELC. The reference voltages or Set
Points (SP) are set as 24V at 0s, 29V at 1s and 24V at 2s of interval. The Control Variable (CV) for the
corresponding reference voltages are mentioned graphically. Figure 12 depicts the regulatory response of LQR
for POELC. The load is increased and decreased by 20% of the interval of 1s.
 Renjini G.S, Deepa THANGAVELUSAMY / GU J Sci, 36(1): x-x (2023)

5. THE RESEARCH FINDINGS AND DISCUSSION

Simulations are done for the evaluation of POELC with PIC and LQR using MATLAB. Simulink tool. Proposed
POELC is designed for regulating the output voltage of 24V for 200W SMPS application. The tuning procedure
of PIC, LQR is already explained in the previous sections.

Figure 11. Servo Response of LQR for POELC

Figure 12. Regulatory Response of LQR for POELC

In the case of a practical system, feedback control loops in closed-loop systems have time delays. Therefore,
the fourth-order system represented in Equation (16) is approximated to First Order system with Dead-
time(FOD). The corresponding output to control input transfer function is obtained as,

𝑉𝑜 (𝑠) 𝐾𝑒 −Ʈ𝑠 (31)

𝑉𝑖𝑛 (𝑠)
= (1+𝑇𝑠)

𝑉𝑜 (𝑠) 2.4𝑒 −0.000004𝑠

=
𝑉𝑖𝑛 (𝑠) (1 + 0.28𝑠)
(32)

where K- system gain, Ʈ – dead time, 𝑇 – time constant.

 Renjini G.S, Deepa THANGAVELUSAMY / GU J Sci, 36(1): x-x (2023)

Figure 13. Servo Response of PIC and LQR for POELC

Figure 13 illustrates the servo response of PIC and LQR in POELC. The reference voltage of the system is
changed from 24V to 29V at 1s and 29V to 24V at 2s. The proposed optimal LQR almost follows the
reference values and it shows excellent servo response in both the first-order and fourth-order systems.
Figure 14 indicates the regulatory response of PIC and LQR in POELC with load disturbance introduced
at t=1s and t=2s. The disturbance is also shown for characterizing the performance of the system with these
two topologies. The proposed LQR controller can achieve the set-point very smoothly even after the load
change has occurred for both FOD and higher-order systems.

Figure 14. Regulatory Response of PIC and LQR for POELC

Table 2. Time Domain Specification of PIC and LQR for POELC

Transient Rise time Peak time Peak Overshoot Settling time
Response tr(ms) tp(ms) Mp(%) ts(s)
PIC 1.5 1 80 0.82
LQR 0.0002 0.000006 0.5 0.000004
 Renjini G.S, Deepa THANGAVELUSAMY / GU J Sci, 36(1): x-x (2023)

Table 2 observes time domain specifications for POELC with PIC and LQR control strategies. Rise time, peak
time, maximum peak overshoot and settling times are evaluated for both controller topologies. LQR shows better
time response specifications. Table 3 illustrates the performance indices for POELC using PIC and LQR. The
value of Integral Absolute Error (IAE) is reduced from 5 with PIC and 0.3 with LQR in the servo region.
Similarly, the value of IAE is reduced from 7 with PIC and 0.5 with LQR in the disturbance region. The
minimum steady-state error implies that the system response is better with LQR techniques. Comparing these
results, it was observed that LQR enhances the system with the stable controlled operation and it is efficiently
used for 200W SMPS applications.

Table 3. Performance Indices of PIC and LQR for POELC

Performance Indices PIC LQR
IAE 5 0.3
ITAE 17.5 1.5
Servo Response
ISE 7 0.1
ITSE 20 0.4
IAE 7 0.5
ITAE 20 1.4
Regulatory Response
ISE 8 0.19
ITSE 20 0.02

6. CONCLUSION

POELC operated in boost mode has been designed for SMPS applications. This system is evaluated in
terms of two controller topologies such as PIC and LQR. System performance is successfully improved by
using LQR with and without delay. By using feed-forward and feedback control techniques in LQR, the
system open-loop poles are investigated and it moves towards the LHP. Results obtained reveals that the
ISE and IAE values are lower with LQR than PIC. The results concluded that the designed converter is best
suitable for SMPS applications.

CONFLICTS OF INTEREST

No conflict of interest was declared by the authors.

REFERENCES

[1] Achiammal, B., Thillaikarasi, M., “Performance Analysis of PI controller for power electronic
Converter usinng evolutionary algorithm”, International Journal of Advance Research in
Science and Engineering, 7(2): 1132–1140, (2018).

[2] Luo, F., L., Ye, H., Rashid, M., H., “Multiple-quadrant Luo-converters”, IEE Proceedings-
Electric Power Applications, 149(1): 9–18, (2002).

[3] Luo, F., L., Ye, H., “Positive output cascade boost converters”, IEE Proceedings- Electric Power
Applications, 151: 590–606, (2004).

[4] Muhammed Abuljaleel Ibrahim., “Performance evaluation of PI controller for positive output
Luo converter”, International Journal Power Electronics and Drive System, 11: 1816-1825,
(2020).

[5] Nouri, T., Hosseini, S., H., Babaei, E., Ebrahimi, J., “A non-isolated three- phase high step-up
DC–DC converter suitable for renewable energy systems”, Electric Power Systems Research, 140:
209–224, (2016).
 Renjini G.S, Deepa THANGAVELUSAMY / GU J Sci, 36(1): x-x (2023)

[6] Kaouane, M., Boukhelifa, A., Cheriti, A., “Regulated output voltage double switch Buck-Boost
converter for photovoltaic energy application”, International Journal of Hydrogen Energy, 41(45):
20847–20857, (2016).

[7] Ramash Kumar, K., Jeevananthan, S., “PI control for positive output elementary super lift luo
converter”, World Academy of Science, Engineering and Technology, 39(3): 732–737, (2009).

[8] Rebeiro, R., S., Uddin, M., N., “Performance analysis of an FLC-based online adaptation of both
hysteresis and PI controllers for IPMSM drive”, IEEE Transactions on Industry Applications,
48(1): 12–19, (2011).

[9] Singh, S., Sharma, D., K., Kishore, K., Botre, B., A., Akbar, S., A., “Modeling, Simulation, and
Implementation of Fast Settling Switched PI Controller for MOX Integrated Pt Micro heater”,
IEEE Sensors Journal, 18(20): 8549–8557, (2018).

[10] Renwal, D., Kumar, M., “Hybrid PI-fuzzy logic controller based DC-DC converter”, Proceedings
of the 2015 International Conference on Green Computing and Internet of Things, ICGCIoT 2015,
1: 753–757, (2016).

[11] Chincholkar, S., H., Chan, C., Y., “Comparative study of current-mode controllers for the positive
output elementary Luo converter via state-space and frequency response approaches”, IET Power
Electronics, 8(7): 1137–1145, (2015).

[12] Jazi, H., N., Goudarzian, A., Pourbagher, R., Derakhshandeh, S., Y., “PI and PWM Sliding Mode
Control of POESLL Converter”, IEEE Transactions on Aerospace and Electronic Systems, 53(5):
2167–2177, (2017).

[13] Mahdavi, M., Shahriari-Kahkeshi, M. and Abjadi, N. R., “An Adaptive Estimator-Based Sliding
Mode Control Scheme for Uncertain POESLL Converter”, IEEE Transactions on Aerospace and
Electronic Systems, 55(6), 3551–3560, (2019).

[14] Usharani, V., Agasthiya, R., “Enhanced MPPT Technique For DC-DC Luo Converter Using
Model Predictive Control For Photovoltaic Systems”, International Journal of Engineering
Research and Development, 11(01), 2278–67, (2015).

[15] Wang, B., Kanamarlapudi, V., R., K., Xian, L., Peng, X., Tan, K., T., So, P., L., “Model Predictive
Voltage Control for Single-Inductor Multiple-Output DC- DC Converter with Reduced Cross
Regulation”, IEEE Transactions on Industrial Electronics, 63(7): 4187–4197, (2016).

[16] Montagner, V., F., and Fabricio, H., D., “A Robust LQR Applied To A Boost Converter
With Response Optimized Using A Genetic Algorithm”, XVIII Brazilian Congress of Automatics,
2297–2302, (2010).

[17] Abdullah, M., A., Tan, C., W., and Yatim, A., H., “A simulation comparison of PI and Linear
Quadratic Regulator controllers in DC-DC converter”, IEEE Conference on Energy Conversion,
1: 37–41, (2015).

[18] Mohammed Ahsan, A., d., M., Federico, M., “Modelling and Simulation of PI-controllers
Limiters for the Dynamic Analysis of VSC-based Devices”, IEEE Transactions on Power
Systems, 8950(c): 1–9, (2019).

[19] Luo, F.L., and Ye, H., “Advanced DC / DC Converters”, 3rd Edition, CRC Press, Boca Raton,
USA, (2004).
 Renjini G.S, Deepa THANGAVELUSAMY / GU J Sci, 36(1): x-x (2023)

[20] Mohan, N., Underland, T., Robbins, W., “Power Electronics - Converters, Applications and
Design”, 2nd Edition, John Wiley and Sons Inc, USA, (2003).

[21] Lindiya, A., Palani, S., Iyyappan, M.,“Performance comparison of various controllers for DC-DC
synchronous buck converter”, Procedia Engineering, 38: 2679– 2693, (2012).

[22] Guo, L., Hung, J., Y., Nelms, R., M., “Comparative evaluation of sliding mode fuzzy controller
and PID controller for a boost converter”, Electric Power Systems Research, 81(1): 99–106,
(2011).

[23] Paraskevopoulos, P., N., “Modern control engineering”, 1st Edition, CRC Press, (2001).

[24] Ogata, K., “Modern control engineering”, 5th Edition, Prentice Hall, (2017).