为了提高接地网故障诊断的准确性,提出一种基于信息熵与证据融合的接地网故障诊断方法.该法通过高频激励获取可及节点的电压信号,根据信息熵理论分别提取信号的时域特征奇异谱熵、频域特征功率谱熵、时频域特征小波包能谱熵,并将这3个特征量作为证据体,运用证据理论进行多特征信息融合,然后将融合结果输入经误差修正算法改进后的径向基函数神经网络中,实现对接地网的故障识别.仿真实例表明,与由单一特征量得到的诊断结果相比,此方法故障诊断的可靠性和准确性更好.%In order to improve the accuracy of fault diagnosis of grounding grids,a fault diagnosis method for grounding grids is proposed based on information entropy and evidence fusion.In this method,the voltage signals at touchable nodes can be obtained under high-frequency excitation.According to the information entropy theory,the time domain characteristic singular spectrum entropy,frequency domain characteristic power spectrum entropy and time-frequency domain characteristic wavelet packet energy spectrum entropy are extracted respectively,which are further taken as evi-dence body.Multi-feature information fusion is carried out by using evidence theory,then the fusion result is input to a radial basis function(RBF)neural network improved by an error correction algorithm to implement the fault identifica-tion of grounding grids.Simulation results show that compared with the diagnosis results obtained by a single feature, the reliability and accuracy of fault diagnosis by using the proposed method is better.
展开▼
