基于VMD样本熵和AFSA优化ELM的配电网单相接地故障选线方法

Single-Phase Grounding Fault Line Selection Method for Distribution Network Based on VMD Sample Entropy and AFSA Optimized ELM

针对目前配电网单相接地故障选线存在的问题,提出一种基于变分模态分解(Variational modal decomposition,VMD)样本熵和人工鱼群算法(Artificial fish swarm algorithm,AFSA)优化极限学习机(Extreme learning machine,ELM)的故障选线新方法。该方法首先采集线路故障后一个周期的暂态电流信号并进行VMD分解,再计算各条线路的样本熵并形成故障特征向量,利用AFSA优化ELM的输入权值与隐含层阈值来建立AFSA-ELM模型,最后将故障特征向量分别输入到ELM和AFSA-ELM模型中进行训练和测试得出选线结果。Matlab仿真对其可行性和有效性进行验证,结果表明:文中所提出的方法有效地提高了选线的速度与精度。

In view of the existing problems in single-phase grounding fault line selection in distribution network,a new fault selection method based on variational mode decomposition(VMD)sample entropy,artificial fish swarm algorithm(AFSA)and optimized extreme learning machine(ELM)is proposed.Firstly,the transient current signal of one period after line fault is collected and decomposed by VMD.Then,the sample entropy of each line is calculated and the fault eigenvector is formed.The AFSA-ELM model is established by optimizing the input weights and hidden layer thresholds of ELM by AFSA.Finally,the fault feature vectors are input into the ELM and AFSA-ELM models respectively for training and testing to obtain the results of line selection.The feasibility and validity of the proposed method are verified by MATLAB simulation.The results show that the proposed method effectively improves the speed and accuracy of line selection.