International Journal of Computer Applications (0975 – 8887) Volume 56– No.2, October 2012 1 Optimal Tuning of PID Controller for DC Motor using Bio-Inspired Algorithms Nitish Katal Student Department of ECE, ASET Amity University, Rajasthan, India. Sanjay Kr. Singh Associate Professor Dept. of ECE, Anand International College of Engineering, Jaipur, Rajasthan, India. ABSTRACT This paper presents the performance comparison between the various soft computing techniques used for optimization of the PID controllers, implemented for speed control system for a DC motor. PID controllers are extensively used in industrial control because of their simplicity and robustness, but when industrial control is imperilled by external glitches, leads to the instability of the system. PID controller optimization using soft-computing algorithms lays emphases on obtaining the best possible PID parameters for improving the stability of the system. The PID controller has been implemented for speed control of a DC motor and the results obtained from optimization using soft-computing are compared with the ones derived from the Ziegler-Nichols method, and comparatively better results are obtained in Stimulated Annealing case. Keywords PID Controllers, Controller Optimization, DC Motor, Genetic Algorithms, Stimulated Annealing, Multi-objective Genetic Algorithms 1. INTRODUCTION Since there is a tremendous development in the power electronic systems, yet the direct current machines are the prime source for the generation of the electric traction. Now a days, finding more useful applications in automobile industry in case of electric vehicles. Since, in DC motors, by adjusting the terminal voltage we can operate it over a wide range of speeds, thus making them compatible with most mechanical loads by virtue of their torque/speed characteristics, thus delivering high performance and easy controllability [1, 2]. But, in real time applications, there are certain factors like external noise, variable and uncertain inputs, unknown parameters, changes in the dynamics of the load, etc.; leading to the instability in their control. PID controllers cause of their simplicity and robustness finds applications in 90% of the control systems in use today. So, the optimization of the PID controller parameters is one of the most important fields in implementation and designing of PID controllers. The classical and widely accepted method for tuning the PID parameters is computation by Ziegler-Nichols [3] method. However, computing the gains doesn’t always provides the best parameters because tuning criterion presumes one-fourth reduction in the first two peaks [4]. But in real time applications, because of the noise, the tuned parameters does not always give the best results, so need is there to even fine tune them, so that they can easily adapt with these changing system dynamics. For better adaptive response of the system, in presence of external glitches, the use of various soft computing techniques like Fuzzy-Logic, Artificial Neural Networks, Genetic Algorithms, Particle Swarm Optimization, Neuro Fuzzy, Neuro-Genetic, etc. have ceded better results. In this paper, the optimization of the PID controller gains has been carried out using Genetic Algorithms (GA), Multi- Objective Genetic Algorithms (Mobj-GA) and Stimulated Annealing, while using the Ziegler-Nichols parameters for the determination of the lower and upper bound limits for the initialization of PID parameter populations. Then, the optimization of the PID controllers for the estimation of the best PID parameters has been done with respect to the objective function, stated as, “Sum of the integral of the squared error and the squared controller output deviated from its steady-state” According to the results obtained in this paper, considerably better results have been obtained in the case of the Stimulated Annealing, when compared to the other techniques in respect of the step response of the system. 2. PID Controllers Proportional Integral and Derivative –PID controllers because of their simplicity and wide acceptability, are playing an imperative role in control systems, and for regulating the closed loop response in industrial controls, PID controllers alone contribute 90% of all the PID’s used today. A PID controller based system is represented in simple block level diagram as in Figure 1. Figure 1. Schematic representation of unity feedback PID controller system architecture. The general equation for a PID controller for the above figure can be given as [5]:     dt s dR K dt s R K s R K s C d i p ) ( ) ( ) ( . ) ( Where K p , K i and K d are the controller gains, C(s) is output signal, R(s) is the difference between the desired output and output obtained. Some of the prime methods for tuning are: Mathematical criteria, Cohen-Coon Method, Trail and Error Method, Ziegler-Nichols Method and now a days the Soft-Computing