GA-based constrained optimal control of a hydraulic two-tank switched system

This work suggests a hybrid approach combined with a conventional method and a metaheuristic to study the problem of constrained optimal control of nonlinear switched systems. The first method is based on the Pontryagin Maximum Principle associated with the Lagrange multipliers method while satisfying the KKT conditions. Concerning the second method, it uses the Genetic Algorithms. Finally, this approach is illustrated through a hydraulic two-tank system application.

I. Introduction

Switched systems have been the interest of various researchers as they make part of different real processes that would not be depicted by only continuous or only discrete models, like manufacturing, communication networks, automotive and chemical processes [1]–[3]. These systems stand for a category of hybrid systems that are simply depicted through piecewise dynamics, periodically illustrated within a specified period. In every interval, the system is in a different and unique mode, determined by a control rule [4]–[6].