Genetic algorithm tuned pid controllers for multivariable level control system

Abstract

In this paper, we have optimised the parameters of Proportional Integral and Derivative (PID) controllers using Genetic Algorithm (GA). The PID controllers were integrated with coupled tank system. This coupled tank system is a Multi Input Multi Output (MIMO) system which also depicts a nonlinear behaviour. GA was implemented using GA framework in DEAP library. Operations and parameters in GA were selected in order to obtain globally optimized parameters for the PID controllers and to avoid the prematurity of the population. The system function was developed considering the nonlinear behaviour of the system and PID parameters were obtained based on this model equation. The performances of the GA tuned controllers were compared with respect to the Zeigler Nicholson (ZN) method in terms of Integral of the Squared Error (ISE) values, rising time and overshoot.