Research On Gas Turbine Generator Excitation PID Control System Based On Rough Set Neural Network

The gas generator excitation control system is of great significance to the gas turbine power system operation, for it can not only ensure the voltage quality of the whole system and distribute reactive power on each grid unit, but also can help to improve the stability and economy of the power system. This paper proposes a new excitation control method, which applies variable precision rough control and the radial basis function neural network technology to nonlinear control design for gas generator.This thesis first elaborates the advantages and disadvantages of all kinds of excitation controller design method, and deduces nonlinear mathematical model of gas power generation single machine infinite system based on the deep analysis of the gas generator and its excitation control system principle, and then establishes the VPRS-RBF neural network by blending the variable precision rough control and RBF neural network, and achieves the intelligent control of automatic adjustment of PID excitation controller parameters according to the VPRS-RBF neural network control rule through the rough identification. Finally, the thesis takes a large number of simulation tests on VPRS-RBF neural network PID excitation controller, and compares it with the conventional PID excitation controller.The result shows that the design of excitation controller has stronger robustness and adaptability.