小波分解结合PSO-DBN的高压直流输电线路单极接地故障测距

Single-pole Grounding Fault Location of HVDC Transmission Lines Based on Wavelet Decomposition and PSO-DBN

针对已有故障定位方法的不足,提出了一种小波分解结合深度信念网络DBN的故障定位方法。对不同故障距离、不同过渡电阻下的线模、零模电压进行分析,推导出故障距离与小波分解各尺度模极大值之比的非线性关系。以小波分解模极大值之比作为故障特征,在MATLAB平台上搭建DBN故障定位模型用粒子群优化算法对模型参数优化,仿真结果表明所提方法测距精度较高,过渡电阻耐受能力较强。

Aiming at the shortcomings of existing fault location methods,a fault location method based on wavelet decomposition and deep belief network(DBN)is proposed.The line mode and zero mode voltages under different fault distances and transition resistors are analyzed,and the nonlinear relationship between the failure distance and the ratio of the modulus maximum of each scale of wavelet decomposition is derived.I aking the ratio of wavelet decomposition modulus maxima as the fault feature,a DBN fault location model is built on the MATLAB plat form and the model parameters are optimized by particle swarm optimization algorithm.The simulation results show that the proposed method has high ranging accuracy and strong transition resistance tolerance.