Performance evaluation of swarm intelligence on model-based PID tuning

Abstract

PID controller has been implemented in many applications due to its simplicity and its good performance. The main problem in PID controller design is tuning its parameters. In order to result in optimal performance, PID parameters should be tuned precisely. An alternative approach that can be used in PID parameters tuning is using swarm intelligence including Particle Swarm Optimization (PSO) and Artificial Bee Colony Optimization (ABCO). This paper presents the performance evaluation of both techniques on PID controller tuning. The tuning is done offline based on a model of plant. The objective function is minimizing the mean square error of step response. Both techniques result in the same optimal solution and produce better response characteristics compared to conventional PID tuning by Ziegler-Nichols method and manual tuning.