火电厂生产过程故障诊断神经网络专家系统

将人工神经网络和专家系统结合地火电厂生产过程的故障诊断，提出并设计了热工过程故障诊断系统。同时对常规ＢＰ算法做了改进，神经网络则采用分层化与并联运用相结合的结构，从而使系统诊断速度和准确率得到提高。仿真结果表明了该系统设计的合理性和可行性。

Fault Diagnosis Based on Neural Network and Expert System for a Thermal Power Plant

The paper is about the study of controlling the very complex working process in thermal power plant, the study and design of the fault diagnosis system adopts a method of combining the neural network with expert system. Based on dividing groups, the general structure of neural network is presented, in which the method of parallel connections are simplified by reducing unnecessary weight distributions. The pace auto - adjusted algorithm is introduced into BP network. In addition, outer -combine is replaced by inter -com-bine of the neural network and expert system. So the fault diagnosis system has a high calculation rate and fault diagnosis precision. The simulation results show the design of the fault diagnosis system is reasonable and feasible.