基于粒子群模糊神经网络的继电保护系统故障诊断

Relay Protection System Fault Diagnosis Based on Particle Group Fuzzy Neural Network

针对电力系统继电保护中故障诊断的特点和要求,建立了基于模糊神经网络的故障智能诊断系统模型。采取粒子群优化(PSO)算法和误差反向传播(BP)算法相结合的方法训练该模型网络,充分发挥PSO全局寻优能力和BP局部细致搜索优势,提高了诊断的可靠性和准确性。实验结果证明了该方法的有效性。

Aiming at the characteristics and requirements of the fault diagnosis in power system relay protection system, fault intelligent diagnosis system model based on fuzzy neural network was established. The method of particle swarm optimization （PSO） algorithm combined with error back propagation （BP） algorithm was used to train the model network, bring full play of PSO all round optimizing ability and BP partial optimizing superiority, to raise the reliability and correctness of the diagnosis. Experimental results prove the validity of the method.