基于混合智能算法的直流输电线路故障测距方法

摘要

A fault location method based on hybrid intelligent algorithm for high voltage direct current (HVDC) transmission line is proposed. It corrects the results of fault location method based on genetic algorithm by the way of neural network algorithm, with the ranging accuracy improved when there are faults at two terminals of transmission line. Inheriting the advantages of fault location method based on genetic algorithm, the proposed method can locate faults with two terminals’ voltage and current of line in the time domain and is free from the effect of fault position and transition resistance. What’s more, the influence on ranging accuracy by line parameters deviation is greatly reduced. The proposed method is simulated by PSCAD and MATLAB, the simulation results show that the proposed method is valid and the combination of neural network algorithm and genetic algorithm can locate the fault on HVDC transmission line accurately and quickly.%提出了一种基于混合智能算法的直流输电线路故障测距方法。利用神经网络算法对基于遗传算法的故障测距方法在线路两端故障时的测距结果进行了修正，提高了线路两端故障的测距精度。该算法利用故障后线路双端的电压、电流量，在时域内进行故障测距，继承了基于遗传算法的故障测距方法的优点，不受故障点位置与过渡电阻的影响，且测距结果精度受线路参数偏差的影响较小。应用PSCAD仿真软件及Matlab对所提算法进行仿真验证，仿真结果表明，神经网络与遗传算法相结合，可以实现直流输电线路全线范围内的准确故障测距。