基于模糊神经网络的故障类型识别

提出了基于模糊神经网络的双端电源输电线路故障类型识别的方法,用ATP提取输电线路故障后一周后继电保护安装点的三相电压电流以及反映接地故障的零序电流基频分量及其相应的相角,并采用T-S模型与改进BP算法结合的模糊神经网络,实现故障类型识别。该方法不受故障位置、故障电阻及对两端电源初始相角差、系统运行方式等不确定的因素影响,仿真结果表明该类型识别方法可靠、正确。

Discrimination of the fault types based on fuzzy neural network

The discrimination method of the fault types based on fuzzy neural network of the transmission line in double sources is presented.The technique extracts fundamental component of the three-phase voltages and currents and zero-sequence current that can judge the grounding fault on relaying point after a cycle of fault,adopts fuzzy neural network that is combined by T-S model and improved BP algorithm to discriminate the fault type.It can't be influenced by those uncertain factors including fault location,fault resistance,angle initial difference of double sources and operation modes.Simulation result indicates the technique of fault type discrimination is very reliable and accurate.