Application of GA-FNN hybrid control system for hydroelectric generating units

A kind of hybridized control system based on genetic algorithms and fuzzy neural networks is presented to control hydroelectric generating units in the paper. The presented system combines basic fuzzy neural networks controller and optimizing part. Takagi-Sugeno type model is used in fuzzy neural networks. Designed genetic algorithms have the character of adaptive optimization techniques and are used to optimize the parameters and rules of fuzzy neural network controller. In order not to impact the control performance of fuzzy neural networks controller in optimizing process, RBF identifying networks, knowledge base and virtual fuzzy neural networks controller are employed to accomplish together the work of optimizing and training on-line. Simulation results show that the hybridized control system is feasible and stable, and the controlling performance of the hybridized system is superior to conventional fuzzy controller or PID controller.