RBF网络在电力电子电路故障诊断中的应用

Application of the RBF network in fault diagnosis for power electronic circuit

电力电子电路模型具有很强的非线性,通常对其进行故障诊断比较困难。而神经网络具有非线性．自适应性、并行性和窖错性等．故可由神经网络来学习和存储电力电子主回路的故障信息和故障类型(或原因)之间的映射关系,并将其用于诊断,从而实现电力电子电路的故障诊断。

Model of power electronic circuit has severely nonlinearity and it has a great difficulty to diagnose its malfunction usually. However, neural network has nonlinearity, self-adaptability, parallelity, fault tolerant ability and so on, so it can be used to learn and store mapping relations between fault information of power electronic circuit and fault types, and to diagnose fault, then fault diagnosis of the power electronic circuit will be implemented.