基于LSTM算法的线路故障智能诊断方法研究

Research on Intelligent Diagnosis Method for Line Faults based on LSTM Algorithm

传统识别算法对复杂电力线路短路故障的特征提取较为困难,为了进一步地提高线路故障诊断的准确率,提出了一种新的故障诊断方法用于对线路短路故障的类型进行辨识。该方法基于长短时记忆(Long Short-Term Memory,LSTM)深度网络模型,直接将线路故障后的三相故障电流采样序列作为模型的输入,通过迭代训练建立与故障类型之间的非线性映射关系。通过仿真及实例验证表明,所提方法在不同工况条件下,对线路的10种短路故障类型可以达到100%的识别正确率,并且不受故障位置、过渡电阻和系统两端电源相位差等因素的影响。

Traditional recognition algorithms are difficult to extract features of complex power line short circuit faults.In order to further improve the accuracy of line fault diagnosis,a new fault diagnosis method is proposed for identifying the types of line short circuit faults.The method is based on the long short-term memory(LSTM)deep network model.The three-phase fault current sampling sequence after the line fault is directly taken as the input of the model,and the nonlinear mapping relationship between the fault type and the fault is established through iterative training.Through simulation and example verification,it is shown that the proposed method can achieve 100%recognition accuracy for 10 types of short circuit faults on the line under different operating conditions,and it is not affected by factors such as fault location,transition resistance,and phase difference between the power supplies at both ends of the system.