PSG COLLEGE OF TECHNOLOGY

(Autonomous Institution)

COIMBATORE – 641 004

A report submitted in partial fulfillment of the requirements for the course

OPTIMAL CONTROL (18UC27)

LQR AND LQG CONTROLLER DESIGN ON CONTROL OF

QUADRUPLE TANK PROCESS
KOMALADEVI N (18MO05)
I Year (Second Semester)

Programme: M.E. Control Systems

Branch: DEPARTMENT OF INSTRUMENTATION AND CONTROL SYSTEMS
ENGINEERING
PSG COLLEGE OF TECHNOLOGY
COIMBATORE – 641 004

1
 TABLE OF CONTENTS
CHAPTER Page No
1. INTRODUCTION .………………………………………..........................................................3
2. CONTROL OF QUADRUPLE TANK PROCESS……………………………………………...4
3. EXPERIMENTAL SETUP...…………………………………………………….........................6
3.1. Aim …….…………………..…….…………………………………………………...........6
3.2. Requirements ….………………………………………………………………………….6
3.3. Problem Statements.………….…………………………………………………………..6
3.4. Procedure.………………………………………………………………………………….6
4. Conclusion …………………………………………………………………..………….……….11
5. Reference ………………………………………………………………………………………...12

2
 CHAPTER 1
INTRODUCTION
Linear–quadratic–Gaussian (LQG) control problem is one of the most
fundamental optimal control problems. It concerns linear systems driven by additive white
Gaussian noise. The problem is to determine an output feedback law that is optimal in the
sense of minimizing the expected value of a quadratic cost criterion. Output measurements are
assumed to be corrupted by Gaussian noise and the initial state, likewise, is assumed to be a
Gaussian random vector. Under these assumptions an optimal control scheme within the class
of linear control laws can be derived by a completion-of-squares argument. This control law
which is known as the LQG controller is unique and it is simply a combination of a Kalman
filter (a linear–quadratic state estimator (LQE)) together with a linear–quadratic regulator (LQR).
The separation principle states that the state estimator and the state feedback can be designed
independently. LQG control applies to both linear time-invariant systems as well as linear time-
varying systems, and constitutes a linear dynamic feedback control law that is easily computed
and implemented: the LQG controller itself is a dynamic system like the system it controls. Both
systems have the same state dimension. Optimal control is concerned with operating
a dynamic system at minimum cost. The case where the system dynamics are described by a
set of linear differential equations and the cost is described by a quadratic function is called the
LQ problem. One of the main results in the theory is that the solution is provided by the linear–
quadratic regulator (LQR), a feedback controller whose equations are given below. The LQR is
an important part of the solution to the LQG (linear–quadratic–Gaussian) problem. Like the LQR
problem itself, the LQG problem is one of the most fundamental problems in control theory.

3
 CHAPTER 2
CONTROL OF QUADRUPLE TANK PROCESS
Simple processes with only one output that may controlled by one input (variable) are
known as single input single output process. But many processes are not such simple. They
have more than one input (variable) and one output, which are called Multi Input-Multi Output
(MIMO) processes. Common MIMO systems have some difficulties, such that they are large
and complex. In addition, they have nonlinearities and also loop interactions which are between
inputs and outputs. On the purpose of studying multivariable systems and designing controllers,
the quadruple tank process (QTP) is chosen as a benchmark. This system is suitable for
studying linear and nonlinear controllers and exhibits minimum and non-minimum system
behaviors by simply changing configuration valve positions.

Multivariable system involves more than one control loop, These loops interact with each other,
in such a manner that single input not only affects its own output but also affects other process
outputs.[1,2] QTP has four tanks and two pumps, this benchmark is simply a water level control
problem. The aim of the process is to keep the liquid level in the lower tanks at the desired
values. the process is used to show multivariable interactions which are also known as coupling
and it limits performance in multivariable control systems. Multivariable interactions in a QTP
are each output (water levels) of the system has affected by two pumps. it can be regarded as a
prototype for many MIMO control applications in industry such as paper production processes,
chemical processes, metallurgy and biotechnological areas, medical industries.

4
Ai=cross sectional area of the tank
hi= the tank water level
qin_i= in-flow of the tank
qout_i= out-flow of the tank
The inflow of the tank (qin_i) only depends on the input pump
voltage and out-flow of the tank (qout_i) depends on the gravity
and acceleration due to head of the water in the tank.

Control variable

hi-level of the tank

5
 CHAPTER 3
EXPERIMENTAL SETUP
3.1 AIM

To design and estimate the lqr and lqg controller design on control of quadruple tank process.

3.2 REQUIREMENTS

1. State space model

2. Matlab software
3. Pc

3.3 PROBLEM STATEMENT

State space model

6
3.4 PROCEDURE

LQG AND LQR DESIGN

OPEN LOOP RESPONSE

clear all;
close all;
A=[-0.016 0 0.044 0;0 -0.011 0 0.033;0 0 -0.044 0;0 0 0 -0.033];
B=[0.0833 0;0 0.0628;0 0.0479;0.0312 0];
C=[0.5 0 0 0; 0 0.5 0 0];
D=0;
sys=ss(A,B,C,D);

%OPEN LOOP ANALYSIS

openpoles=eig(A);
t=0:0.0001:25;
u1=zeros(size(t));
u2=zeros(size(t));
u=[u1 ; u2];
x0=[0.5 0 1.5 1];
[y,t,x]=lsim(sys,u,t,x0);
plot(t,y);
co=rank(ctrb(A,B));
ob=rank(obsv(A,C));

%pole placements
p1=-10-10i;
p2=-10+10i;
p3=-50;
p4=-57;

%pole placement controller design

K=place(A,B,[p1,p2,p3,p4]);
sys_cl=ss(A-B*K,B,C,D);

CLOSED LOOP RESPONSE

%CLOSED LOOP IMPULSE RESPONSE

% u1=zeros(size(t));
% u2=zeros(size(t));
% u=[u1 ; u2];
x0=[0.5 0 1.5 1];
[yp,t,xp]=lsim(sys_cl,u,t,x0);
plot(t,yp);

7
LQR CONTROLLER DESIGN
%lqr controller design
Q=C'*C;
R=eye(2);
K1=lqr(A,B,Q,R);
sys_cl2=ss(A-B*K1,B,C,D);

%CLOSED LOOP IMPULSE RESPONSE

x0=[0.5 0 1.5 1];
[yp2,t,xp2]=lsim(sys_cl2,u,t,x0);
plot(t,yp2);
ci2=-K1*xp2';
ci=-K*xp';

8
 LQG OBSERVER DESIGN

%LQG observer design

ob1=-15+1i;
ob2=-15-1i;
ob3=-15;
ob4=-20;
L=place(A',C',[ob1 ob2 ob3 ob4])';
At1=[A-B*K1 B*K1
zeros(size(A)) A-L*C];
Bt1=[B
zeros(size(B))];
Ct1=[C zeros(size(C))];
sysob=ss(At1, Bt1, Ct1, 0);
[yp2,t,xp2]=lsim(sysob,u,t,[x0 x0]);
plot(t,yp2);

%plot true and estimated state variables

n=4;
e=xp2(:,n+1:end);%error eqn
xo=xp2(:,1:n);%actual value
xo_est=xo-e;

s1=xo(:,1);%for first variable alone

s1_est=xo_est(:,1);
plot(t,s1,'-r',t,s1_est,':b');

s2=xo(:,2);
s2_set=xo_est(:,2);
plot(t,s2,'-r',t,s2_est,':b');

s3=xo(:,3);
s3_set=xo_est(:,3);
plot(t,s3,'-r',t,s3_est,':b');

s4=xo(:,4);
s4_set=xo_est(:,4);
plot(t,s4,'-r',t,s4_est,':b');

9
OPEN LOOP ANALYSIS

Fig 2:open loop response

In this quadruple tank process having the two inputs open loop response obtained between
level of the tank and time response

CLOSED LOOP IMPULSE RESPONSE

Fig 3: closed loop response

In this closed loop response obtained the poles placements and response of the input one and
input two

10
LQR CONTROLLER DESIGN

Fig 4 : lqr controller design

In this lqr controller design based on the true value estimated value should be controlled

LQG OBSERVER DESIGN

For State 1:

11
 CHAPTER 4
CONCLUSION
The linear system model is obtained and the conditions which the system behaves as
minimum or non minimum phase are given. Minimum phase system structure is chosen for this
study than a centralized augmented state feedback controller is designed. For this controller,
any further performance improvement is not considered. We only show that such a controller
design is possible. Since interactions between levels of tanks are in acceptable range a
decentralized control system design can be another option. In this case, interactions between
tanks may be considered as disturbance. Performance of controllers is compared via
simulations.

12
 CHAPTER 5
REFERENCES
[1] Serkin, Özkan, Mehmet Arıcı, Tolgay Kara,” Modelling, Simulation And Control Of Quadruple
Tank Process” Electrical and electronics Engineering Department, Gaziantep University,
Gaziantep, Turkey

[2] D. A. Vijula, K. Anu, P. M. Honey, P. S. Poorna, “Mathematical Modelling of Quadruple Tank

System”International Journal of Emerging Technology and Advanced Engineering, vol.3, Issue
12, December 2013.

[3] K. H. Johansson, “The quadruple-tank process: a multivariable laboratory process with an

adjustable zero,”
IEEE Trans. Control Syst.Technol., vol.8, no.3, pp.456-
465, May. 2000.

13