Application of improved neural network optimizing PID control in hydroelectric generating unit

Because hydroelectric generating unit system is a non-linear, varying with time and high steps system, conventional PID controller cannot get better controlling performance in the control of hydroelectric generating unit system. In order to overcome the shortcoming of conventional controller, neural network is used to optimize PID control parameters and identify the character of hydroelectric generating unit system in this paper. In the design, improved network structure and learning ways are engaged in training network and identifying network. The designed neural networks have fast convergence and shorten training time. Simulation results show that the designed neural network optimizing PID controller can control hydroelectric generating unit system effectively and the controlling performance is superior to traditional PID, which show that the designed controller is feasible.