基于时域微分的地铁直流牵引系统故障定位

Fault Location Research of DC Railway Traction System Based on Time-domain Differential

快速、准确的故障定位是地铁牵引供电系统急需解决的问题。深入研究了地铁直流牵引系统的结构、参数、运行及故障特征,提出一种基于最小二乘法求解的双端时域微分方程故障定位算法,不受过渡电阻影响。针对地铁变电站间不具备通信条件的情况,进一步提出一种改进的基于遗传算法求解的单端故障定位算法,通过构造优化函数求解故障位置。根据实际运行的某地铁供电系统结构参数进行建模,通过故障仿真对算法进行分析,验证算法的可行性,并利用实测数据进行相关比对,证明了算法的实用性和准确性。

Developing fast and accurate fault location methods for DC railway traction power supply system is an urgent problem. This paper deeply studies the structure, parameters, operation, and fault characteristics of the metro DC traction system, and proposes a time domain differential equation method of double side based on the least square method, which will not be affected by transition resistance. Considering that there are no communication conditions between the subway substations, an improved single-end fault location algorithm based on the genetic algorithm has been put forward. The fault location is found through constructing the optimization function. Further, the models based on an actual subway are built, and the fault location algorithms are analyzed through fault simulations. The results demonstrate the feasibility. Finally, by using the measured data for related comparison, the method proves to be practical and accurate.