基于IELM算法的配电网故障区段定位
Fault Section Location for Distribution Network Based on Improved Electromagnetism-like Mechanism Algorithm
-
摘要: 针对传统智能优化算法在配电网故障区段定位中存在准确率随配电网规模扩大而大幅下降的问题,提出一种基于改进仿电磁学(IELM)算法的配电网故障区段定位方法。首先,基于分层处理以及全局寻优思想对传统故障定位方法的一体式结构进行优化,通过改变算法结构实现单一故障及多重故障的分层处理并缩减多重故障定位的解空间规模。在结构优化的基础上对仿电磁学(ELM)算法进行改进,忽略全局电荷对单一电荷的影响,突出单一电荷对最优电荷的学习能力,提高其全局寻优能力及运算效率。最后,分别以IEEE 13节点单电源辐射配电网和改进后的4电源119节点配电网为仿真测试系统进行了测试。测试结果表明,所提故障区段定位方法在面对不同结构和规模的配电网时具有准确性高、容错性好、定位速度快以及降维效果佳等优点,在大规模配电网中具有良好的应用前景。Abstract: Aiming at the problem that the accuracy of traditional intelligent optimization algorithms in fault section location of distribution networks will greatly decrease with the expansion of the scale of distribution networks, a fault section location method for distribution networks based on improved electromagnetism-like mechanism(IELM) algorithm is proposed. Firstly, the integrated structure of the traditional fault location method is optimized based on the idea of hierarchical processing and global optimization. By changing the algorithm structure, the hierarchical processing of single-fault and multiple-fault is realized, and the solution space of multiple-fault location is reduced. Based on the structure optimization, the electromagnetism-like mechanism(ELM) is improved. Therefore, the impact of the global electric charge on the single electric charge is ignored; the learning ability of the single electric charge to the optimal electric charge is highlighted; and the global optimization ability and operation efficiency are improved. Finally, the IEEE 13-bus radiation distribution network with single power source and the modified 119-bus distribution network with four power sources are employed as simulation test systems. The test results demonstrate that the proposed fault section location method has the advantages of high accuracy, good fault tolerance, fast location speed and good dimension reduction effect for the distribution networks with different structures and scales and it also has a good application prospect in large-scale distribution networks.