高阶小波包奇异谱熵在故障选线中的应用研究

摘要

Aiming at the problem that the insensitivity of singular entropy to high resistance grounding faults often causes the failure of selecting fault line, the paper proposes a method of multi-resolution singular spectrum entropy based on higher-order cumulant.This method organically integrates the advantages of the theory of wavelet packet analysis, singular value decomposition theory and information entropy theory, and fully reflects the intrinsic signal information, effectively extracts components of nonlinear characteristics and suppresses Gaussian noise, and then it contains more information after the higher-order cumulating.The ATP-based simulation results show that the fault line selection method can quickly identify the fault line, is less susceptible to the failure time, transition resistance, fault location and other factors, so it has a good adaptability.%针对奇异熵对经高阻接地的不敏感性、经常导致选线失败的问题,提出了基于高阶累积量多分辨率奇异谱熵的方法.该方法有机地融合了小波包分解理论,奇异值分解理论和信息熵理论的各自优点,并充分体现信号的内在信息,有效地提取非线性特征分量和抑制高斯噪声,经高阶累计后包含了更多的选线信息.基于 ATP 的仿真结果表明:该故障选线方法能够快速识别故障线路,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.