基于神经网络的水轮机调节模型辨识

The Identification on Turbine Regulating System Based on Neural Network

针对复杂的水轮机调节系统辨识问题,首先分析水轮机调速器特性,建立通用水轮机调速器模型,并采用PRBS信号仿真其动态特性,进而基于神经网络理论,开展水轮机调速系统辨识分析。研究表明,辨识得到的参数与待辨识参数的误差较小,网络输出与实际系统输出的差值也较小,因此用BP神经网络进行水轮机调节模型辨识是可行的。

Pointed to identification problems of turbine regulating system , the regulating system char-acteristics was analyzed to establish the mathematic model , and using the PRBS signals to simulated the regulating process , and then the identification analysis is carred out based on neural network.The research shows that the error between identified parameters and actual parameters is small , and the error between simulated output and actual output is small , too.So the turbine regulating system iden-tification based on neural network is feasible.