Novel Method on Using Evolutionary Algorithms for PD Optimal Tuning
This paper presents a method to solve the Evolutionary Algorithm (EA) problems for optimal tuning of the Proportional-Deferential (PD) controller parameters. The major efficiency of the proposed method is the Genetic Algorithm (GA) stuck avoidance as well an appropriate estimation for GA lower and upper bounds. Also by this method for the Particle Swarm Optimization (PSO) methodology the initial choice of the controller parameters can be fulfilled to achieve the acceptable performance accuracies. For both GA and PSO methods, the Linear Quadratic Regulator (LQR) obtained trend is used as the reference for the determination of the aforementioned bounds and initial guess. The presented algorithm was applied to regulate a PD controller for the attitude control of a virtual satellite and also with Hardware-in-the-loop (HIL) reaction wheels. Heavy burden trying and error was eliminated from the PD controller design which can be mentioned as the important merit of the presented study.