PSO APPLIED TO DESIGN OPTIMAL PD CONTROL FOR A UNICYCLE MOBILE ROBOT

International Journal of Advance Robotics & Expert Systems (JARES) Vol.1, No.1,
33
PSO APPLIED TO DESIGN OPTIMAL PD CONTROL
FOR A UNICYCLE MOBILE ROBOT
Sarra Bencharef and Hamid Boubertakh
Department of Automatic, Automatic Laboratory of Jijel (LAJ)
University MSBY of Jijel, Algeria
ABSTRACT
In this work, we propose a Particle Swarm Optimization (PSO) to design Proportional Derivative
controllers (PD) for the control of Unicycle Mobile Robot. To stabilize and drive the robot precisely with
the predefined trajectory, a decentralized control structure is adopted where four PD controllers are used.
Their parameters are given simultaneously by the proposed algorithm (PSO). The performance of the
system from its desired behavior is quantified by an objective function (SE). Simulation results are
presented to show the efficiency of the method. ).
The results are very conclusive and satisfactory in terms of stability and trajectory tracking of unicycle
mobile robot
KEYWORDS
Particle Swarm Optimization (PSO), Proportional Derivative, Unicycle mobile robot, Optimization,
Control
1. INTRODUCTION
Today mobile robots are a big hit on the side of the automatic control researchers, for their
various applications in civilian and military tasks.
The control objective of a Unicycle robot can be position stabilization, trajectory tracking…
Various controllers have been developed and adopted by researchers on controlling the
differential drive robot, predictive control and fuzzy logic control [1] , A dynamic-model-based
control scheme for balancing and velocity control of a unicycle robot [2], Fuzzy logic control
(FLC) strategy [3-4], Based on both the kinematic and dynamic model of a unicycle an inner loop
nonlinear controller with a dynamic model controller on the outer loop [5] , back-stepping
technique and the LaSalle's invariance principle [6], adaptive controller [7] , An intelligent
control approach was presented by [8] for a differential drive robot were a higher level navigation
controller combined with a lower level fuzzy logic based controller, feedback linearization with
PD control [9].
The PID control is relevant, due to the simplicity of this structure and its ease of implementation.
However, it has limited performance because of its structure and the linear system and an
opposite non-linear and complex system. In addition, an incorrect setting of these parameters will
lead to a more performance degradation. Unfortunately, there are no systematic methods for
setting parameters and the existing methods do not guarantee optimal operation such as Ziegler
Nichols method [10].
In recent years, many heuristic evolutionary optimization algorithms have been developed. These
include Genetic Algorithm (GA) [11-12], Ant Colony optimization (ACO) [13-14], Particle
Swarm Optimization (PSO) [15], and Gravitational Search algorithm (GSA) [16]. These methods
have acquired a special interest in solving the problem of design of PD controllers.

34
The above listed methods are used for tuning PID controllers in case of SISO systems. In this
paper, we suggest the use of PSO to solve the PD control design problem for control of Unicycle
robot.
2. THE UNICYCLE MOBILE ROBOT
The robot considered in this paper is a unicycle mobile robot, reported by [17], and is described
by the following equation of motion. A basic configuration of Unicycle mobile robot is shown in
Figure 1.
Figure 1. Unicycle mobile robot configuration
Let 𝒒 = 𝒙, 𝒚, 𝜽 be the generalized coordinates of a unicycle, where 𝒙, 𝒚, is the Cartesian
position of the unicycle in a world frame and 𝜽 is its orientation with respect to the 𝒙 axis. The
kinematic model of the system is,
Where:
 𝒗 is linear velocity of the robot.
 𝒘 is the angular velocity of the robot.
The right and left wheel motor velocities can be expressed as 𝒗𝒍 = 𝒓 𝝓𝒍 = 𝝆 − 𝑳 𝒘
and 𝒗 𝑹 = 𝒓 𝝓 𝑹 = 𝝆 + 𝑳 𝒘 respectively.

35
Table 1. The unicycle mobile robot model parameter
3. THE PD CONTROL DESIGN USING
3.1. OVERVIEW OF THE PARTICLE SWARM OPTIMISATION
PSO is nature-inspired heuristic optimization method which first proposed by Kennedy and
Eberhart [15]. It belongs to the category of Swarm Intelligence methods. Its development was
based on mimicking the movement of individuals within a swarm (i.e., fishes, birds, and insects)
in an effort to find the optima in the problem space. It has been noticed that members of the
swarm seem to share information among them. This communication fact leads to increase
efficiency of the swarm. The PSO algorithm searches in parallel using a group of individuals
similar to other population-based heuristic optimization techniques. PSO technique conducts
search using a population of particles, corresponding to individuals. Each particle represents
a candidate solution to the problem at hand. In a PSO system, particles change their positions by
“flying” around in a multidimensional search space. Particle in a swarm adjust its position in
search space using its present velocity, own previous experience, and that of neighboring
particles.
Therefore, a particle makes use of best position encountered by itself and that of its neighbors to
steer toward an optimal solution. The performance of each particle is measured using a predefined
fitness function, which quantifies the performance of the optimization problem. The mathematical
expressions for velocity and position updates are given by.
Where:
𝟏 ≤ 𝒊 ≤ 𝑰 , 𝟏 ≤ 𝒋 ≤ 𝑱 , 𝟏 ≤ 𝒌 ≤ 𝑲
𝑰 . 𝑱 and 𝑲 are respectively the number of particles in the swarm, the dimension of particle, and
the maximum iterations.
𝒙 𝒊𝒋
𝒌
is the position of particle 𝒊 in the dimension 𝒋 at iteration 𝒌
𝒗 𝒊𝒋
𝒌
is the velocity of particle 𝒊 in the dimension 𝒋 at iteration 𝒌 , 𝒑𝒃𝒆𝒔𝒕𝒊𝒋 is a personal best of
particle 𝒊 in the dimension 𝒋, 𝒈𝒃𝒆𝒔𝒕𝒋 is a global best of all particles in the dimension 𝒋, 𝒘 is inertia
weight factor , 𝒄 𝟏 and 𝒄 𝟐 are acceleration constants, 𝒓 𝟏 and 𝒓 𝟐 are random numbers in interval
𝟎 ⋅ 𝟏 .

36
3.2. THE PD PARAMETERS TUNING PROBLEM
The general function of Proportional Integral Derivative controllers (PDs) are given by:
The error 𝒆 𝒕 is defined as:
Where :
𝒖 is the control signal, 𝒚 𝒓 is the reference signal and 𝒚 is the controller output
The design parameters’ vector is
Without loss of generality, a squared error cost function (7) is used to quantify the effectiveness of
a given PD controllers; it is evaluated at the end of desired input signal of the closed-loop system
under control.
In the PD controller design problem using PSO, a particle is one value of the PD controller
parameters vector.
A particle swarm is a group of parameters vector values which are manually or randomly initialized
in search space defined by limits of the components of the tuning parameters vector. So, a the
particle position 𝒙 is the PD controller parameter value vector, 𝑷𝒃𝒆𝒔𝒕 the best parameters values
vector regarding the cost function, 𝑮𝒃𝒆𝒔𝒕 is the global best parameters values vector of all
particles.
4. APPLICATION TO THE CONTROL FOR THE UNICYCLE MOBILE ROBOT
The control objective is to control the mobile robot to achieve a desired point. That is, we use two
decentralized PD controllers .
The design parameters’ vector is composed of the parameters of the PD controllers:
The effectiveness of the PD controllers is evaluated at the end of the response of the entire control
system to desired input signals by the squared error fitness function given by:
𝑰𝑺𝑬 = 𝒆(𝒕) 𝟐
𝒅𝒕
𝒕 𝒇
𝒕 𝟎
𝟐
𝒊=𝟏 (𝟗)
In the simulations, we take the:
 The initial states : 𝒙 𝟎 = 𝟎 𝒎 , 𝒚(𝟎) = 𝟎 𝒎.

37
 The desired point has the following coordinates 𝒙 𝒕 𝒇 = 𝟏 𝒎 , 𝒚(𝒕 𝒇) = 𝟏 𝒎
 The cost function spans from 𝒕 𝟎 = 𝟎 𝒔 , 𝒕 𝒇 = 𝟐𝟓 𝒔
 For the PSO we take the parameters in Table 2.
 The initial parameters values are chosen randomly between the bounds given by Table 3.
Table 2. The algorithme parameters
Table 3. The initial parameters bounds
4. SIMULATION RESULTS
The final parameters values are given on Table 4.
The Figure 2 shows the evolution of the cost function with iterations. Figures 3, 4 and 5 show
respectively the position in the x-axis, the position in the y-axis and the position in the plane with
the tuned controllers.
Table 4. The final tuned parameters values

38
Figure 2. The Cost function
Figure 3. The position in the Y-axis
Figure 4. The position in the X-axis

39
Figure 5. XY plot
From these figures, we can see well a good tracking of the desired values of the position in the X-
axis and Y-axis and on the XY plane for the algorithms PSO Algorithm, (illustrated respectively
by Fig. 3.4and 5).
After the results we can see that the proposed algorithm give us satisfactory results because it
provide good trajectory tracking of unicycle mobile robot. The robot was able to follow its own
way, we can simply say that PSO Algorithm has shown its effectiveness in terms of stability of
unicycle mobile robot.
5. CONCLUSION
This paper has presented the performance of a intelligent synthesis of PD controller, using the
praticle swarms optimisation algorithme (PSO) to control a unicycle robot . The intelligent
controller was implemented and evaluated by simulating the system model in MATLAB
environment. The controller gains have been optimised using PSO optimisation algorithm to
minimise the mean square error of the robot system. The performance of the system with the
optimised controller gains has been greatly improved proving a successful control and optimisation
strategy. The simulation was able to yield results that showed good stability of the unicycle robot.
The simulation results show the efficiency of the proposed methode.
REFERENCES
[1] X.Jiang; Motai, Y.; Zhu, X., (2005), Predictive fuzzy logic controller for trajectory tracking of a
mobile robot, Soft Computing in Industrial Applications, Proceedings of the 2005 IEEE Mid-Summer
Workshop on , pp.29-32.
[2] Han, S.I., Lee, J.M., (2015), Balancing and Velocity Control of a Unicycle Robot Based on the
Dynamic Model, Industrial Electronics, IEEE Transactions on , vol.62, no.1, pp.405-413
[3] Castillo, O., Martínez-Marroquín, R., Melin, P., Valdez, F., & Soria, J., (2012), Comparative study of
bio-inspired algorithms applied to the optimization of type-1 and type-2 fuzzy controllers for an
autonomous mobile robot. Information Sciences, 192, pp.19-38
[4] Martínez, R., Castillo, O., & Aguilar, L. T., (2009), Optimization of interval type-2 fuzzy logic
controllers for a perturbed autonomous wheeled mobile robot using genetic algorithms. Information
Sciences, 179(13), 2158-2174.

40
[5] R.Carona, A.P.Aguiar, J.Gaspar, (2008), Control of unicycle type robots tracking, path following and
point stabilization.
[6] Ti-Chung Lee, Kai-Tai Song, Ching-Hung Lee, and Ching-Cheng Teng, (2001), Tracking control of
unicycle-modeled mobile robots using a saturation feedback controller. IEEE Transactions on Control
Systems Technology, 9(2), pp.305-318.
[7] Martins, F. N., Celeste, W. C., Carelli, R., Sarcinelli-Filho, M., & Bastos-Filho, T. F. (2008). An
adaptive dynamic controller for autonomous mobile robot trajectory tracking. Control Engineering
Practice, 16(11), pp.1354-1363.
[8] M.F.R.Lee, F. H. S.Chiu, (2013), Intelligent multi-behavior control and coordination for the
autonomous differential drive mobile robot, In IEEE International Conference on Fuzzy Theory and
Its Applications (iFUZZY),pp. 31-36.
[9] De Luca, A., Oriolo, G., & Vendittelli, M. (2000, September). Stabilization of the unicycle via
dynamic feedback linearization. In 6th IFAC Symp. on Robot Control (pp. 397-402).
[10] K. J. Åström, T. Hägglund, PID Controllers, Theory, Design and Tuning, Instrument Society of
America, Research Triangle Park, North Carolina, 1995.
[11] J. M. Adams, and K. S. Rattan, « A genetic multi-stage fuzzy PID controller with a fuzzy switch, » In
Proc. IEEE Int.Conf. on Systems, Man, and Cybernetics, Tucson, AZ, USA, 4,2001, pp. 2239-2244.
[12] C.-N. Ko, T.-L. Lee, H.-T. Fan, and C.-J. Wu, « Genetic Auto-Tuning and Rule Reduction of Fuzzy
PID Controllers, » Proc. IEEE Int. Conf. on Systems. Man, and Cybernetics,Taipei, Taiwan, 2006, pp.
1096-1101.
[13] I. Pan, S. Das, and A. Gupta, « Tuning of an optimal fuzzy PID controller with stochastic algorithms
for networked control systems with random time delay», ISA Transactions , Vol. 50, No 1, pp. 28-36,
2011.
[14] H. Boubertakh, , M. Tadjine, , P.-Y. Glorennec, S. Labiod, « Tuning fuzzy PID controllers using ant
colony optimization». In: Proc. of the 17th Mediterranean Conference on Control and Automation
(MED'09), Thessaloniki, Greece, 2009, pp. 13–18.
[15] J. Kennedy and R. Eberhart, “Particle swarm optimization,” Proc IEEE Int Conf on Neural Networks,
pp. 1942-1948, 1995.
[16] Esmat Rashedi, Hossein Nezamabadi-pour, Saeid Saryazdi, « GSA: A Gravitational Search Algorithm
»,Department of Electrical Engineering, Shahid Bahonar University of Kerman, P.O. Box 76169-133,
Kerman, Iran.
[17] Klancar, G., Matko, D., Blazic, S., (2005), Mobile robot control on a reference path. In Intelligent
Control, Proceedings of the 2005 IEEE International Symposium on, Mediterrean Conference on
Control and Automation (pp. 1343-1348).
AUTHORS
Sarra Bencharef is currently a Ph.D. student at the Automatic Control Department of Jijel University,
Algeria. She received the Engineer degree in Electronic from Jijel University, Algeria, in 2009, and the
Magister degree of Automatic Control and signal processing at the Automatic Control Department of Jijel
University, Algeria, in April 2014. She is with Automatic Laboratory of Jijel (LAJ). His research interests
include Aerial robotics, Linear and Nonlinear control, intelligent control and optimization
Hamid Boubertakh graduated from National Polytechnic school of Algiers (ENP) in Algérie in 2009. He
currently works as a Teacher-Researcher at the University of Jijel, Algeria, where he continues his research
activities. His areas of interest are intelligent control and optimization.

PSO APPLIED TO DESIGN OPTIMAL PD CONTROL FOR A UNICYCLE MOBILE ROBOT

More Related Content

What's hot (20)

Similar to PSO APPLIED TO DESIGN OPTIMAL PD CONTROL FOR A UNICYCLE MOBILE ROBOT (20)

More from JaresJournal (20)

Recently uploaded (20)

PSO APPLIED TO DESIGN OPTIMAL PD CONTROL FOR A UNICYCLE MOBILE ROBOT