Report On

Robust Sliding-Mode Control Design

for a Voltage Regulated Quadratic
Boost Converter

Under the supervision of

Prof. Somnath Maity
By
Baijayanti Mala Das
215EE3143
CONTENTS
SL. No PG. No.

1 Introduction 2

2 Literature 2

3 Objective 3

4 Circuit Operation 4

5 Sliding mode analysis 5

6 Stability Analysis 9

7 Result 12

8 References 13

1|Page
INTRODUCTION:
The need of developing an ad hoc electronic equipment for power distribution systems, electric
vehicles and energy backup architectures which are supplied by batteries, fuel cells, photovoltaic
modules or other sources or other dc sources is prompting engineers to use 400v dc bus as a core
of the distribution. Although both transformer based and transformer less topologies can be
proposed ,transformer less topologies has advantage over transformer based because of power
density, weight, size and cost .The single switch converter derived from conventional boost
converter(such as quadratic and cubic) gives better solutions as they operate within safe duty
cycle avoiding modular saturation. Quadratic boost converter shows a better tradeoff between
efficiency and duty cycle. Moreover, robust control of the boost converter has become an
attractive research field in past years due to the non minimum phase nature of its control to
output transfer function, which results from the linearization of the converter averaged equations
around the steady state and whose parameters have a high degree of uncertainty.

LITERATURE REVIEW

The sliding mode control is used in this approach to design a hysteresis based quadratic boost
converter that provides a regulated output voltage of 400V dc from an input in the range of 15-
25V dc. In this paper, the definition of a simple sliding surface for the regulation of the input
inductor current yields the indirect control of the output voltage by forcing the mentioned current
to reach a desired reference value in the equilibrium state. Therefore, if the current reference in
the sliding surface is modified by the action of a PI compensator processing the output voltage
error, it will be possible to regulate the output voltage to a desired level. Hence the proposed
controller consists of two loops, namely, an inner loop for input inductor current control and an
outer loop establishing the reference for the inner loop to ensure the output voltage regulation.
Hence, to cope with the parameter uncertainty, a robust loop shaping method is chosen to
synthesize the PI compensator with robustness constraints. The values of PI gains Kp and Ki are
obtained for the stable range given by the Routh-Hurwitz stability test using a geometrical
analysis is in the Nyquist diagram involving the maximization of the integral gain. This Robust
Loop Shaping synthesis technique will be referred in the following as RLSMIGO.

2|Page
OBJECTIVE:

 To design a robust controller to obtain output voltage regulation in a quadratic boost

converter with high dc gain.

 The controller has an inner loop based on sliding-mode control whose sliding surface is
defined for the input inductor current.

 The current reference value of the sliding surface is modified by a proportional-integral

compensator in an outer loop that operates over the output voltage error.

 To prove the stability of the two-loop controller using the Routh– Hurwitz criterion,
which determines a region in the Kp −Ki plane, where the closed-loop system is always
stable.

 To perform the analysis of the sliding-mode-based control loop by means of the

equivalent control method.

 To derive the outer loop compensator by means of the Nyquist-based Robust Loop
Shaping approach with the M-constrained Integral Gain Maximization technique.

QUADRATIC BOOST CONVERTER CIRCUIT OPERATION:

Mode -1

when the controlled switch is in ON state, diodes D1 and D3 are open and diode D2 is closed,
inductor L1 takes energy from the input power supply and inductor L2 takes energy from
capacitor C1. At the same time, output capacitor C2 gives the energy to the load.

Mode -2

when the controlled switch is in OFF state, diodes D1 and D3 are closed and diode D2 is open,
inductors L1 and L2 charge capacitors C1 and C2 and give the energy demanded by the load.
Therefore, the converter operates continuously between ON and OFF states.

3|Page
CIRCUIT DIAGRAM:

L1 D2 L2

S
D3 D1
Vi
C1 C2
R

CIRCUIT EQUATIONS:

diL1 vi vC1
   1 u 
dt L1 L1
diL 2 vC1 vC 2
  1 u
dt L2 L2
dvC1 i i
  L 2  L1  1  u 
dt C1 C1
dvC 2 v i
  C 2  L2  1  u 
dt RC2 C2

STEADY STATE ANALYSIS

Since in steady state operation the waveforms must repeat from one time period to next, the
integral of the inductor voltage VL1 and VL2 over one time period must be zero

TS

v
0
L1 0

TS

v
0
L2 0

The expansion of the above two equation gives voltage gain equation as shown in Equation

4|Page
vo vC 2 1
 
vin vC1  1  D  2

At steady state the net charge in the capacitor current is zero.

TS
IO
i
0
C1 0 I L1 
 1 D
2

TS
IO
i
0
C2 0 I L2 
 1 D
PARAMETERS VALUES

Power rating 100W

Input voltage 20V

Output voltage 400V

Switching frequency 100KHZ

Input inductor 1 125uF

Input inductor 2 4.2mH

Capacitor 1 8uF

Capacitor 2 9uF

SLIDING MODE ANALYSIS

STEP 1: Applying the equivalent control method to obtain ueq.
STEP 2: Obtaining the ideal sliding-mode dynamics by replacing the equivalent control.
STEP 3: Obtaining the equilibrium point.
STEP 4: Linearization of the ideal sliding dynamics around the equilibrium point.
STEP 5: Obtaining the characteristic polynomial.
STEP 6: Determination of the stability conditions by using the Routh-Hurwitz test.

5|Page
LINEAR MODELING OF THE IDEAL SLIDING DYNAMICS
Sliding surface which is defined using the input inductor current

s ( x )  iL1  I E (t )
where
t
I E (t )   K p  vC 2 (t )  vref   Ki   vC 2 ( )  vref  d 


The switching function u yielding sliding motions is expressed as follows:

1 , when s ( x ) <0
u=
{
0 , when s ( x ) > 0

Evaluating the existence condition S(x)·d(S(x))/dt < 0

the local reachability condition

dI E (t )
0  vi  L1  vC1
dt
Applying the invariance conditions

S(x)=0and d(S(x))/dt=0
1 dIE
ueq=1−
vc 1 (
vi−L 1
dt )

Ideal sliding dynamics of the converter

diL 2 (t ) vC1 vC 2  dI E (t ) 
   vi  L1 
dt L2 L2 vC1  dt 
dvC1 i i  dI (t ) 
  L 2  L1  vi  L1 E 
dt C1 C1vC1  dt 
dvC 2 v i  dI (t ) 
  C 2  L 2  vi  L1 E 
dt RC2 C2 vC1  dt 

6|Page
Obtaining equilibrium points considering ,

,
By solving the equivalent control equations , equilibrium points can be obtained as:
3 1
vC1  vi4  I E R  4
1
vC 2   vi I E R  2
1
3
 v 4
iL 2  I  i 
4
E
R

Linearizing around the above equilibrium points the system equations can be obtained as

diL 2 (t ) 2vC1 vi vC 2 L1vC 2 dIE (t ) vC 2vi

   
dt L2 L2 vC1 L2 vC1 dt L2vC1
dvC1 i i v v I L i dIE (t ) iE vi
  L 2  E C1  i E  1 E 
dt C1 C1vC 2 C1vC1 C1vC1 dt C1vC1
dvC 2 v v i i v L i dI (t ) i v
  C 2  i L 2  L 2 C1  1 L 2 E  L 2 i
dt RC2 C2 vC1 C2vC 2 C2vC1 dt C2 vC1

Obtaining A,B,C,D parameters:

A( s)VC 2 ( s)  B ( s) I E ( s )  C ( s)Vi ( s)  D( s) I O ( s)
 C    C T   4v
A( s)  L2C2 vC 2 s 3  TL 2vC 2  K m2 2  1 s 2  vC 2  2 2  K m2 L 2   vi  s  C 2
 C1    C1 TC1   TC1
K m2 R Rv
B ( s)  K m2 TL 2TL1 RvC 2 s 3  TL1 RvC 2 s 2   2TL1vC 2  TL 2vi  s  2 i
TC1 TC1
2 K m2 vC 2
C ( s)  K T v s  vC 2 s 
2
m L2 C 2
2

TC1
K m2 TL 2 RvC 2 2 RvC 2
D( s )  TL 2 RvC 2 s 2  s
TC1 TC1
vC 2 vC1 L L
Km   , TL1  1 , TL 2  2 , TC1  RC1 , TC 2  RC2
vC1 vi R R

7|Page
Synthesis of voltage regulation loop:
DETERMINATION OF A LOCAL STABILITY REGION
• The PI compensator in the outer loop is defined by the transfer function

• The loop-gain transfer function :

 The closed- loop transfer function from reference voltage to output voltage is given by

Gvo( s)=¿ ¿

 The characteristic polynomial is given by: ¿ where

MAIN CIRCUIT

OPEN LOOP SIMULATION RESULTS:

8|Page
OUTPUT:

INPUT INDUCTOR CURRENT VS TIME OUTPUT VOLTAGE AND LOAD VS TIME RESP
RESP. CURRENT VS

9|Page
Result: Stability region for PI compensator

Graph Equations
po  b0 K i
p1  a0  b1 K i  b0 K p
p2  a1  b2 K i  b1 K p
p3  a2  b3 K i  b2 K p
p4  a3  b3 K p
p0  0
p1  0
p2  0
p3  0
p4  0
p2 p3  p4 p1  0
p1 p2 p3  p4 p12  p0 p32  0

10 | P a g e
CLOSED LOOP SIMULATIONS:

OUTPUT VOLTAGE OUTPUT POWER

11 | P a g e
REFERENCES:

[1] O. Lopez-Santos, L. Martinez-Salamero, G. Garcia, H. Valderrama-Blavi, D.O. Mercuri,

"Efficiency analysis of a sliding-mode controlled quadratic boost converter", IET Power
Electron., vol. 6, no. 2, pp. 364-373, Feb. 2013.
[2] K. J. Aström, H. Panagopoulus, T. Hägglund, "Design of PI controllers based on non-
convex optimization", Automatica, vol. 34, no. 5, pp. 585-601, 1998
[3] Z. Chen, "PI and sliding mode control of a Cuk converter", IEEE Trans. Power Electron.,
vol. 27, no. 8, pp. 3695-3703, Aug. 2012.
[4] H. Panagopoulos, K. Astrom, " PID control design and H \$_{infty}\$ loop shaping
", Proc. Int. Conf. Control Appl., pp. 103-108, Aug. 1999.
[5] J. A. Morales-Saldaña, R. Galarza-Quirino, J. Leyva-Ramos, E. E. Carbajal-Gutierrez, M.
G. Ortiz-Lopez, "Multiloop controller design for a quadratic boost converter", IET Elect.
Power Appl., vol. 1, no. 3, pp. 362-367, May 2007.
[6] J. A. Morales-Saldaña, R. Loera-Palomo, E. Palacios-Hernandez, J. L. Gonzáz-Martínez,
"Modelling and control of a DC-DC quadratic boost converter with R2P2", IET Power
Electron., vol. 7, no. 1, pp. 11-22, Jan. 2014
[7] S.-C. Tan, Y. M. Lai, C. K. Tse, "General design issues of sliding-mode controllers in
DC–DC converters", IEEE Trans. Ind. Electron., vol. 55, no. 3, pp. 1160-1174, Mar.
2008.
[8] R. W. Erickson, D. Maksimovic, Fundamentals of Power Electronics, 2000, Kluwer.
[9] H. K. Khalil, Nonlinear Systems, 1992, MacMillan.

[10] F. Golnaraghi, B.C. Kuo, Automatic Control Systems, 2010, Wiley.

12 | P a g e