基于稀疏分解与综合测度值的配电网故障选线方法

发布时间：2018-11-12 20:58

【摘要】：为提高基于暂态量选线方法的准确性和鲁棒性,同时尽可能多的包含故障信息,提出了一种基于稀疏分解与综合测度值的配电网故障选线方法。首先,利用匹配追踪算法对故障瞬间2个周期内的暂态零序电流进行4次迭代分解;然后,选取1/4周期内的最佳原子作为研究对象,一方面计算原子奇异熵,另一方面利用最佳原子重构得到特征原子,并计算其实际能量;最后,以归一化后原子奇异熵的倒数为自变量,构造了故障信任度函数,进而构造了能够多角度量化故障信息的故障综合测度函数,最大故障综合测度值对应的线路被选取为故障线路。仿真结果表明:故障发生时,求得的故障线路综合测度值大于健全线路综合测度值,基于该特性得到的选线结果不受故障线路类型、故障相角、过渡电阻和故障距离的影响。
[Abstract]:In order to improve the accuracy and robustness of the transient line selection method and to include as much fault information as possible, a fault line selection method based on sparse decomposition and synthetic measure is proposed. Firstly, the transient zero-sequence current in two cycles is decomposed four times by matching tracking algorithm. Then, the best atom in 1 / 4 cycle is selected as the research object. On the one hand, the singular entropy of atom is calculated; on the other hand, the characteristic atom is obtained by the best atom reconstruction, and its actual energy is calculated. Finally, taking the reciprocal of the normalized atomic singular entropy as the independent variable, the fault trust function is constructed, and the fault synthesis measure function which can quantify the fault information from multiple angles is constructed. The line corresponding to the maximum fault synthesis value is selected as the fault line. The simulation results show that the comprehensive measure value of fault line is larger than that of sound line when fault occurs, and the line selection result based on this characteristic is not affected by fault line type, fault phase angle, transition resistance and fault distance.
【作者单位】：中国农业大学信息与电气工程学院;西安交通大学电气工程学院;河南理工大学电气工程与自动化学院;
【基金】：国家自然科学基金青年基金(51407186;61403127)~~
【分类号】：TM862

【相似文献】