International Journal of Computer and Electrical Engineering, Vol. 2, No.

1, February, 2010
1793-8163
41



AbstractIn this paper, the design of servo system is
proposed using the state feedback where the parameters of
feedback gain matrix are tuned using the particle swarm
optimization (PSO). Type-1 servo system is considered as in
most of the Single Input-Multi Output (SIMO) process the
presence of integrator cannot be identified in the plant. It
contains more than one inter-coupled transfer function
between the state vectors of any non-linear system. One of the
performance objectives for controller design is to keep the
error between the controlled output and the set-point as small
as possible. The Performance Index (PI) considered are
Integral Time Absolute Error (ITAE), Integral Square Error
(ISE), Integral Absolute Error (IAE) for optimal control. The
simulations are tabulated in section 4. to analyze the promising
results by the proposed method.

I ndex TermsParticle Swarm Optimization, Performance
Index, SIMO, Servo system.

I. INTRODUCTION
The type 1 servo system has one integrator in the
feed-forward path, and the system will exhibit no steady-state
error in the step response. In this paper, the pole-placement
approach to the design of type 1 servo system has been
carried out. The system is limited to have a scalar control
signal u and a scalar output [1]. The block diagram is shown
in Figure 1.

Fig.1 Basic form of Type-1 Servo System
Hence, the system equations are given by:
u B x A x + = (1)
x C y = (2)

I
k x K u + = (3)
x C r y r = =
.
(4)

Manuscript received March 23, 2009.
Tushar Jain, M.J. Nigam, Indian Institute of Technology Roorkee,
Uttarakhand, India. Emails: tushaar.iitr@gmail.com(Tushar Jain),
mkndnfec@iitr.ernet.in(M.J. Nigam)
where x = state vector of the plant (n-vector)
u = control signal (scalar)

y
= output signal (scalar)
= output of the integrator (state variable)
r = reference input signal
A = nxn constant matrix
B = 1 nx constant matrix
C = xn 1 constant matrix
It is assumed that the plant is completely controllable. The
basic idea of designing the type-1 servo system here is to
stabilize (n+1)th-order regulator system that will bring the
error to zero.
Type-1 servo system can also be designed for the plant
having an integrator, though this has not been covered much
in detail, as in many non-liner systems the presence of
integrator in the SIMO process is difficult to predict before
linearization.
Linear Quadratic Regulator (LQR) [2] is designed for
optimal control of Inverted Pendulum (IP) using (Genetic
Algorithm) GA; the swing-up controller takes approximately
4 seconds to reach the upright position.
II. PARTICLE SWARM INTELLIGENCE
Particle Swarm Optimization (PSO) is swarm intelligence
based algorithm to find a solution to an optimization problem
in a search space, or model and predict social behavior in the
presence of objectives [3]. It is a stochastic, population based
evolutionary computer algorithm for problem solving. Social
influence and social learning enable a person to maintain
cognitive consistency.
The swarm is typically modeled by particles in
multidimensional space that have a position and a velocity.
These particles fly through hyperspace (i.e., ) and have
two essential reasoning capabilities: their memory of their
own best position and knowledge of the global or their
neighborhood's best. For minimizing an optimization
problem, "best" simply means the position with the smallest
objective value. Members of a swarm communicate good
positions to each other and adjust their own position and
velocity based on these good positions. So a particle has the
following information to make a suitable change in its
position and velocity:
1). A global best that is known to all and immediately
updated when a new best position is found by any
particle in the swarm.
Tuning of Type-1 Servo System Using Swarm
Intelligence for SIMO Process
Tushar Jain, Member, IACSIT, IAENG, M.J. Nigam
International Journal of Computer and Electrical Engineering, Vol. 2, No. 1, February, 2010
1793-8163
42

2). Neighborhood best that the particle obtains by
communicating with a subset of the swarm.
3). The local best, which is the best solution that the
particle has seen.
The particle position and velocity update equations [4] in
the simplest form that govern the PSO are given by:
v [] = v [] + c1 * rand () * (pbest [] - present []) + c2 * rand ()
* (gbest [] - present []) (5)
present [] = present [] + v [] (6)
v[] is the particle velocity, present[] is the current particle
(solution). pbest[] and gbest[] are defined as stated before.
rand () is a random number between (0, 1). c1, c2 are
learning factors usually c1 = c2 = 2.
Initialization: the
velocity and place of
particles
Initialization of pbest
and gbest
Evaluation of particles
Velocity updating
fitness value is
better than
pbest
fitness value is
better than
gbest
Set it as the new pbest
Set it as the new gbest
Stopping
Condition met
Stop
Yes
No
Yes
No
No
Yes

Fig.2 Flowchart of PSO
As the swarm iterates, the fitness of the global best
solution improves (decreases for minimization problem). It
could happen that all particles being influenced by the global
best eventually approach the global best, and there on the
fitness never improves despite however many runs the PSO is
iterated thereafter. The particles also move about in the
search space in close proximity to the global best and not
exploring the rest of search space. This phenomenon is called
'convergence'. If the inertial coefficient of the velocity is
small, all particles could slow down until they approach zero
velocity at the global best. The selection of coefficients in the
velocity update equations affects the convergence and the
ability of the swarm to find the optimum. One way to come
out of the situation is to reinitialize the particles positions at
intervals or when convergence is detected. The flowchart of
PSO algorithm is shown in Figure 2.
The five basic principles of swarm intelligence are
articulated [3] as; First, is the proximity principle: the
population should be able to carry out simple space and time
computations. Second is the quality principle: the population
should be able to respond to quality factors in the
environment. Third is the principle of diverse response: the
population should not commit its activities along excessively
narrow channels. Fourth is the principle of stability: the
population should not change its mode of behavior every time
the environment changes. Fifth is the principle of
adaptability: the population must be able to change behavior
mode when its worth the computational price.
It can be viewed as a mid-level form of A-life or
biologically derived algorithm, occupying the space in nature
between evolutionary search, which requires eons, and
neural processing, which occurs on the order of milliseconds.
III. SIMO CASE STUDY
To investigate the impact of PSO on the design of type-1
servo system using state feedback, a nonlinear SIMO process:
Inverted Pendulum system is considered.
A. Inverted Pendulum
The inverted pendulum control problem [5] is usually
presented as a pole balancing task. The system to be
controlled consists of a cart and a rigid pole hinged to the top
of the cart. The cart can move left or right on a
one-dimensional bounded track, whereas the pole can swing
in the vertical plane determined by the track. The linearized
system equations around t u = in the state space are:
( )
(
(
(
(

(
(
(
(
(
(

+ +
+
+ +

+ + + +
+
=
(
(
(
(

u
u
u
u

x
x
Mml m M I
m M mgl
Mml m M I
mlb
Mml m M I
gl m
Mml m M I
b ml I
x
x
0
) (
) (
) (
0
1 0 0 0
0
) ( ) (
0
0 0 1 0
2 2
2
2 2
2
2

u
Mml m M I
ml
Mml m M I
ml I
(
(
(
(
(
(

+ +
+ +
+
+
2
2
2
) (
0
) (
0
(7)
International Journal of Computer and Electrical Engineering, Vol. 2, No. 1, February, 2010
1793-8163
43



u
x
x
y
(

+
(
(
(
(

=
0
0
0 1 0 0
0 0 0 1
u
u

(8)
Where, M = mass of cart = 0.5 kg
m = mass of pendulum = 0.2 kg
b = friction of cart = 0.1 N/m/sec
I = inertia of pendulum = 0.006 kgm
2

l = length of pendulums center = 0.3 m
F = force applied to cart
The state of the system is defined by values of four system
variables: ( u u

, , , x x ) the cart position, cart velocity,
pendulum angle and angular velocity of the pendulum pole,
respectively. Control force is applied to the system to prevent
the pole from falling while keeping the cart within the
specified limits.
Real time hardware of 1-stage Linear Inverted
Pendulum manufactured by GoogolTech, China
is used for simulation.
IV. RESULTS AND DISCUSSION
The open loop step response of the pendulum angle and the
cart position is shown in Figure 3.
It can be seen that a small force on the cart sets the
pendulum angle at 90
0
therefore, to improve the dynamics of
the system; some controller has to be designed. The Simulink
diagram of type-1 servo system using state feedback
controller for IP is shown in Figure 4.
The performance of the state feedback controller can be
judged by the value of its parameters. The PSO algorithm
explained in Section 2. has been used to estimate the
parameters of the feedback gain matrix
Performance Index (PI) is optimized for position of the
cart as in the real system, the length of the apparatus on
which the cart is moving is limited. So, care has been taken to
restrict the motion of cart within the limits. This is analyzed
on the basis of ISE, IAE and ITAE to maintain the pendulum
position at 0
0
for any disturbance given to the cart.

Fig.3 Open loop step response of IP
The simulated results of controller using PSO for different
types of errors are shown in Figure 5, Figure 6, and Figure 7.

Fig.4 Step response for ISE

Fig.5 Step Response for IAE

Fig.6 Step response for ITAE
Neuro-PID controller [6, 7] via error-back-propagation
(BP) method is designed using GA as optimization technique;
it took much time to get convergence results. Also, the
International Journal of Computer and Electrical Engineering, Vol. 2, No. 1, February, 2010
1793-8163
44

overshoot of the pendulum angle is large in comparison to
techniques discussed above.





TABLE 1. FEEDBACK GAIN MATRIX W.R.T. PI
PERFORMANCE
INDEX
W.R.T.
ISE
W.R.T.
IAE
W.R.T.
ITAE
FITNESS VALUE 0.000083 0.01363 0.008278
K1 -35.217 -47.7867 -117.6538
K2 -51.453 -28.3160 -159.5329
K3 17.096 111.3929 51.1659
K4 71.769 57.0990 124.0727
KI -191.633 -702.5304 -1513.3
V. CONCLUSION
For the SIMO process under consideration, type-1 servo
system proves to be a promising controller employing PSO
algorithm as compared to [2], [6]-[7]. Figure 5, 6, and 7
shows the effectiveness of the proposed method for single
stage Linear Inverted Pendulum system. As it is analyzed, a
single PID controller cannot be used with real system and
employing two PID controllers for controlling cart position
and pendulum angle, requires much computation time for
number of iterations.
While designing state feedback controller, the
computation time is less using search algorithm for the
optimal feedback gain matrix. In PSO, the search region is
independent of the boundary condition specified by the
number of parameters to be tuned, it is given by the distance
between the randomly selected initial position and position
corresponding to optimal fitness value.
PSO algorithm combined with other intelligent techniques,
such as neural networks, expert systems, and fuzzy logic
control systems open a new way to design and construct
control systems adapted to complex process.
REFERENCES
[1] Katsuhiko Ogata, Modern Control Engineering, Prentice Hall, New
Jersey, 3
rd
edition-1997
[2] Hamid R. P., M. R. Jaheh-Motlagh, Ali-Akbar J., Optimal feedback
control design using genetic algorithm applied to inverted pendulum,
IEEE International Symposium on Industrial Electronics, pp. 263-268,
June,2007
[3] James Kennedy, Russell Eberhart, Particle Swarm Optimization, IEEE
Transactions, pp. 1942-1948, 1995
[4] Qi Kang, Lie Wang, Qi-di Wu, Research on Fuzzy Adaptive
Optimization Strategy of Particle Swarm Algorithm, International
Journal of Information Technology. pp. 65-77, Vol.12, No.3, 2006
[5] Dadios, Elmer P.; Williams, David J., A Fuzzy-Genetic Controller for
the Flexible Pole-Cart Balancing Problem, Proceedings of IEEE
International Conference on Evolutionary Computation, pp 223, ISBN:
0 7803 2902 3, Nagoya, Japan, 1996.
[6] Sigeru Omatu, Safaai Deris, Stabilization of Inverted Pendulum by the
Genetic Algorithm, Proceedings of IEEE, pp. 700-705, 1996.

Fig.4 Simulink block of Inverted Pendulum