Parameter tuning of PID controller with reactive nature-inspired algorithms

Abstract

A PID controller is an electrical element for reducing the error value between a desired setpoint and an actual measured process variable. The PID controller operates according to its input parameters, which need to be set before its run. The optimal values of these parameters must be found during the so-called tuning process. Today, this process can be automatized using stochastic, nature-inspired, population-based algorithms, such as evolutionary and swarm intelligence-based algorithms. Unfortunately, these algorithms are too time consuming, and so the reactive, nature-inspired algorithms using a limited number of fitness function evaluations are proposed in this paper. Two reactive evolutionary algorithms (differential evolution and genetic algorithm), and four reactive, swarm intelligence-based algorithms (bat, hybrid bat, particle swarm optimization and cuckoo search), were used to tune the PID controller in our comparative study. Only ten individuals and ten iterations (generations) were used in order to select the most appropriate optimization algorithm for fast tuning of controller parameters. The results were compared using statistical analysis and showed that particle swarm optimization is the best option for such a task.