论文部分内容阅读
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10kV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。
Three kinds of wave-front extraction algorithms based on mathematical morphology theory, such as mathematical morphological gradient algorithm, morphological-wavelet synthesis filter algorithm and morphological unsampled wavelet decomposition algorithm, were studied. For a practical 10 kV overhead line-cable hybrid The railway power through the line fault ranging simulation test, comparative analysis of the three algorithms in different fault types, different fault distances, different transition resistance and different noise levels under the conditions of adaptability. The results show that the mathematical morphological gradient algorithm is fast and suitable for low-noise conditions. The morphological-wavelet comprehensive filtering algorithm has high ranging accuracy and is suitable for noise-less conditions. The morphological non-sampling wavelet decomposition algorithm is robust against noise High, suitable for a strong noise environment. Based on this, a comprehensive morphological distance measurement scheme is proposed, which can get more rapid and accurate results than the single method, which provides a basis for the development of subsequent system devices.