泛化的信息熵目前已在数据挖掘、信号识别和故障诊断等领域得到了广泛有效的应用。小波熵是一种典型的泛化熵测度,其在电力系统故障检测和识别中的应用具有巨大的潜力。探索各类小波熵算法的机理及其所揭示的电力系统现象本质是进一步开拓其应用的前提。对于小波熵算法之一的小波奇异熵,该文在研究小波变换、奇异值分解及信息熵理论的基础上,分析了其应用原理和本质。基于对小波奇异熵的研究,将其应用于高压输电线路的故障相识别,并提出了基于故障暂态的故障选相判据和方案。基于PSCAD/EMTDC的故障仿真结果表明:该故障选相方案能快速准确地识别各类故障,并且不易受到故障时刻、过渡电阻、故障位置等因素的影响,具有较好的适应性。
Extended Shannon-Entropy has been widely and effectually used in data-mining, signal-identification and fault-diagnosis; Wavelet-Entropy is one of the typical extended Shannon-Entropy and its application in power fault detection and identification has a good prospect. To research the essence of various Wavelet-Entropies is the precondition of developing their application. As for Wavelet Singular Entropy which is one of the Wavelet-Entropies, the paper analyzed its essence on basis of Wavelet Transform, Singular Value Decomposition and Shannon-Entropy, then applied it into faulty phase selection of HV transmission line and put forward the concrete selecting method which used faulty transient voltage. Simulations in PSCAD/EMTDC show: this method can select the faulty phase rapidly and accurately and it won't be affected by faulty types, faulty time, transition resistance or faulty location so it will have a good adaptability.