Neural Network Control Of The Turbine Governor System

The speed-adjusting system of hydraulic turbine is a device that is in order to adjust the rotation speed and output power through automatically adjusting gate opening according to rotation speed. It can also maintain electricity frequency and the stability of electricity voltage. Speed-adjusting system is a key sub-system in the hydro-electricity power station system, in which PID module is the key module. Adopting NN(Neural Network)-PID controller to replace the original PID controller module can improve the flexibility , stability and robust of the speed-adjusting system.The thesis focuses on the establishment of speed-adjusting system module in the hydro-electricity system, the improvement on PID controlling module of the core module of hydraulic governor with the application of neural network control theory and the comparison between new and old module. This paper first introduces neural network control theory. Secondly it analyses the speed-adjusting system in hydro-electricity power station system, elaborating on the physical module and its theory, mathematical module and operational situation. The thesis adopts modular modeling method to mechanically model the speed-adjusting system, and to itemize sub-systems based on their different functions. Then with the application of neural network control theory, it improves PID module in the speed-adjusting system and replaces its original PID module with NN-PID controller. Last the paper tests and simulates the improved mathematics module and compares the results with those of the old module, including testing and simulation on the stable characteristics and dynamic characteristics, and proves the validity and superiority of the NN-PID module.