中低压配电网故障定位的改进蚁群算法研究

Research on Improved Ant Colony Algorithms for Fault Location in Medium and Low Voltage Distribution Networks

蚁群算法在配电网故障定位中取得了良好的应用效果,但其具有搜索时间长、速度慢等特点,改进的蚁群算法能有效改善这些问题。首先,基于区域划分思想,将中低压配电网分为中压、低压配电网进行故障筛选,分别形成稀疏矩阵,以此完成对蚂蚁行走路线的预分解,降低解空间。然后,进一步利用聚类的思想对故障点进行深度分解,求解全局最优路径,提高寻优概率。并从初始化信息素稀疏矩阵及启发因子改进着手,引导算法跳出局部最优。此外,从工程应用角度,设计了故障定位与用电信息采集等各相关系统间的业务关系框架。试验结果验证了研究方法可极大提升停电事件统计的及时性和准确性。

Ant colony algorithm has achieved good application results in distribution network fault location,but it has the characteristics of long search time and slow speed,The improved algorithm of it can effectively improve these problems. Firstly,based on the idea of regional division,the medium and low voltage distribution network is divided into medium and low voltage distribution network for fault screening,respectively forming a sparse matrix,thus completing the pre-decomposition of the ant walking route and reducing the solution space. Then,using the idea of clustering,the fault points are deeply decomposed,the global optimal path is solved,and the optimization probability is improved. Starting from the initialization of the pheromone sparse matrix and the improvement of the heuristic factor,the algorithm is guided to jump out of the local optimum. In addition,from the perspective of engineering application,designed the business relationship framework,such as power failure fault location and power consumption information collection. The experimental results verify that the research method can greatly improve the timeliness and accuracy of statistics on blackout events.