Charged System Search Algorithms for Optimal Tuning of PI Controllers

This paper suggests the application of Charged System Search (CSS) algorithms to the optimal tuning of PI controllers. The presentation is focused on a class of second-order processes with an integral component and variable parameters. The CSS algorithms solve the optimization problems which aim the minimization of objective functions expressed as the integral of squared control error augmented by the integral of squared output sensitivity functions. The expressions of the output sensitivity functions are derived from the state sensitivity models of the control systems with respect to process gain variations. The optimal tuning of PI controllers is proposed such that to ensure a reduced sensitivity with respect to process gain variations. A case study is included to illustrate the application of the new CSS algorithms and to compare it with a Particle Swarm Optimization and a Gravitational Search Algorithm.