Application of Soft Computing in the Control of Hydroelectric Generating Unit

Abstract

Hydroelectric generating unit system of water plant is a non-linear and complicated system. Conventional controller cannot get good controlling performance in control. In this study, an advanced soft computing technique based on fuzzy, neural network and genetic algorithm is used in the control of hydroelectric generating unit system of hydropower plant. Fuzzy reasoning system is used as controller and genetic algorithm is employed to optimize the parameters and rules of fuzzy controller in the design of controller or real-time control process. Dynamic identification model of control system based on RBF neural networks is designed to appraise the controlling performance of fuzzy controller. In the design, improved genetic algorithm is adopted based on general genetic algorithm and the character of control system. The improved genetic algorithm quickens optimizing speed and makes fuzzy controller acquire knowledge effectively. Simulation results show that the designed soft computing optimization control system can well control hydroelectric generating unit and its control performance is superior to conventional controller.