电力电子电路故障的遗传进化神经网络诊断

Fault Diagnosis of Power Electronic Circuit Based on Genetic Manipulation and Artificial Neural Network

介绍了一种诊断电力电子电路故障的基于改进遗传进化的神经网络混合算法 ,论述了该算法的理论和运算操作步骤后将其用于电力电子三相整流电路的故障诊断。仿真实验证明该法可诊断电力电子三相整流电路故障 ,且收敛速度快 ,诊断准确度高 。

In this paper, a hybrid algorithm that combines improved genetic manipulation with artificial neural network is applied to fault diagnosis of power electronic three-phase rectifier circuits. First, in the paper the hybrid algorithm or theory that improved genetic manipulation and artificial neural network is introduced, then this hybrid fault diagnosis method of power electronic rectifier circuits is applied to fault diagnosis of three-phase rectifier circuit. Simulation diagnosis result of practical examples show that proposed hybrid algorithm is faster and more accurate than other algorithm in the convergence rate and fault diagnosis, the proposed fault diagnosis method fulfils requirements of three-phase rectifier circuits, the hybrid algorithm can also be used for reference in other application in engineering and practice.