采用信息融合技术的电网故障诊断方法

为了实现电网故障的准确诊断,首先提出了电网潮流指纹的相关概念和故障潮流指纹识别算法,并在此基础上结合径向基函数(radial-basis function,RBF)神经网络和信息融合技术,提出了一种新的电网故障诊断方法。该方法通过分析电气量信息实现电网故障诊断,避免了开关量和保护信息的误动、拒动等不确定因素。同时,采用改进的线性整数规划法优化配置同步向量测量单元(phasor measurement unit,PMU),利用PMU实时采集潮流信息,克服了传统数据采集与监控系统(supervisory control and data acquisition,SCADA)数据采集密度低且不同步的缺点,能够把握时间间隔为s级或者ms级的连锁故障本质。仿真算例表明,该方法信息需求量小、准确有效,具有很大的工程实用价值,在电网在线故障诊断领域有很好的应用前景。

Novel Approach of Fault Diagnosis in Power Grid Using Information Fusion

In order to diagnose fault of power grid accurately,we firstly define the electrical power grid flow fingerprint,then put forward a flow fingerprint identification algorithm and a novel method for fault diagnosis in which RBF(radial-basis function)neural network and information fusion technology are applied.The proposed method is able to eliminate the uncertainty resulted from relay or breaker mal-operation and refuse-operation scenarios by utilizing electrical quantities,especially when complex fault occur...