一种基于BP与分层变迁WFPN相融合的输电网故障诊断方法研究

A fault diagnosis method for the transmission grid based on the fusion of BP and hierarchical transitional WFPN

为降低电网故障诊断中因人为主观因素的影响而造成的误差,提出了一种基于BP与分层变迁的加权模糊Petri网(weighted fuzzy Petri net,WFPN)相融合的输电线路故障诊断方法。根据故障信息确定出可疑元件,然后针对各元件分别建立它们的子模型和综合诊断模型。考虑到Petri网模型与BP神经网络在结构和形式上有一定的相似性,因此本文采用BP算法对Petri网模型中的权值进行训练。仿真结果表明该方法具有合理性及有效性。

To reduce the errors caused by the subjective factors in power grid fault diagnosis,a fault diagnosis method for the transmission grid based on the fusion of BP and hierarchical transitional weighted fuzzy Petri net（WFPN）is proposed.Firstly,determine the suspicious components according to the fault information,then built the sub-model and the comprehensive diagnosis model for various components.Considering the Petri net model and the neural network have certain similarities in structure and form,this paper uses BP algorithm to train weights of Petri net.The simulation results prove that thisdiagnosismethod is reasonable and effective.