Research Of Hydrogenerator Governor Intelligent Control

Turbine governor, which ensures the stable function of hydropower generator unit, is important control equipment in hydropower plant. This dissertation focuses on the application of intelligent control in turbine governor.Turbine is an unlinear and unminimum-phase system, and it is difficult to build an accurate model for it. This dissertation analyses the defects of the widely used simplified model, and replaces it with a more accurate unlinear model. Based on the basic dynamic equations of turbine, the unlinear model takes the variation of stable head and no-load flow, which are ignored in simplified model, into consideration, so it approximates the real turbine more closely.Trained with the data from the unlinear model, this dissertation uses neural network to identify turbine. In the training process, the adaptive learning rate and error batch-mode process are introduced to accelerate the training rate. The supervised training alleviates over-matching, reduces the possibility of being locked into a local-minimum area and shortens the convergence period.A N-PID intelligent control strategy, which integrates the PID control into a three-layer neural network, is implemented to intelligently control turbine. The forward calculation of control signal and back-propagation of network-weight modification equations are given, along with the principle to specify the initial net-weight values. The output of the network is set with limit. Simulation shows turbine controlled with the N-PID governor can achieve good static and dynamic characteristic. This dissertation uses Genetic Algorithm to choose optimum parameters for the widely used PID controllers. In the optimization calculation process, mutation rate, cross rate and parameter range are adaptively changed to accelerate optimization process. The fitness function is also changed according to different requirements. Simulation shows the Genetic Algorithm can quickly choose optimum PID parameters with high robustness.Utilizing the internal high-rate counter in FX2N series PLC and simple external circuit, the author real-timely measures the hydro generator frequency with high accuracy and reliability, which meets the engineering application requirements.