【摘要】 针对现有的高压输电线路故障信号分析方法难以有效同时判断出所有线路故障类型，以及故障信号特征丢失且难以反映故障特征的问题，提出一种基于故障暂态电流信号的变分模态分解(VMD)结合粒子群优化支持向量机(PSO-SVM)的高压输电线路故障选相方法.在Matlab/Simulink中对多工况高压输电线路的不同故障参数进行仿真提取，将提取的各相故障信号首先进行变分模态分解并计算；然后将每个分量的包络熵值作为特征向量，带入粒子群优化的支持向量机中进行故障识别.仿真结果表明，VMD结合PSO-SVM的输电线路故障诊断模型精度可高达98%以上，且识别精度不受故障类型影响.更多还原

【Abstract】 Aiming at the existing fault signal analysis methods of high-voltage transmission lines, the known analysis methods are difficult to effectively determine all the types of line faults that often occur at the same time, and the fault signal features are lost and it is difficult to reflect the essential characteristics of fault information, a fault phase selection method for high-voltage transmission lines based on fault transient current signal(VMD) combined with particle swarm optimization support vector machine(PSO-SVM) is proposed. In MATLAB/Simulink, different fault parameters of multi-condition high-voltage transmission lines are simulated and extracted, and the extracted fault signals of each phase are first decomposed in variational mode and the envelope entropy value of each component is calculated as a feature vector, and brought into the support vector machine for particle swarm optimization for fault identification, and the simulation results show that the accuracy of the transmission line fault diagnosis model combined with VMD combined with PSO-SVM can be as high as 98%, and the recognition accuracy is not affected by the fault type.更多还原