基于数学形态学原理的行波波头提取算法在铁路电力贯通线测距中的适应性分析

对3种基于数学形态学原理的行波波头提取算法——数学形态学梯度算法、形态学–小波综合滤波算法和形态学非抽样小波分解算法进行了研究，针对一条实际的10 kV架空线–电缆混合铁路电力贯通线路进行了故障测距仿真试验，对比分析了3种算法在不同故障类型、不同故障距离、不同过渡电阻以及不同噪声水平工况下的适应性。结果表明，数学形态学梯度算法运算速度快，适用于低噪工况；形态学–小波综合滤波算法测距精度高，适用于噪声不大的工况；形态学非抽样小波分解算法噪声耐受性高，适用于有较强噪声干扰的环境。基于此，提出了形态学综合测距方案，该方案可获得比单一方法更为快速准确的结果，为后续系统装置的研制提供了依据。

Adaptability Analysis on Fault Location of Continuous Railway Power Line by Traveling Wave Front Extraction Algorithm Based on Mathematical Morphology

Three traveling wave front extraction algorithms based on mathematical morphology, i.e., mathematical morphological gradient (MMG) algorithm, morphological un-decimated wavelet (MUDW) decomposition algorithm and morphology-wavelet (MW) integrated filtering algorithm are researched. By means of the simulation of fault location for a actual 10kV overhead line-cable hybrid continuous railway power transmission line, the comparative researches on adaptability of the three algorithms under different fault types, different fault distances, different transition resistances and different noise levels are performed. Simulation results show that computation speed of MMG algorithm is fast, so it is suitable to low noise level condition; due to its high accuracy of fault location, the morphology-wavelet integrated filtering algorithm is suitable to the condition with noise not too high; the MUDW decomposition algorithm possesses good tolerance of noise, so it is suitable to the fault location under strong noise circumstance. On this basis, a comprehensive morphological fault location scheme is proposed by which a faster and accurate fault location result can be obtained. The results of this research is available for reference in the development of following-up system devices.