Partition fault location of active distribution network based on AC-DBO

Abstract

Aiming at the problems such as slow convergence speed, precocious convergence and poor fault tolerance performance of dung beetle optimizer (DBO) algorithm in solving active distribution network fault location, a partition fault location method for active distribution network based on adaptive chaotic DBO (AC-DBO) algorithm is put forward. Firstly, an improved Tent chaotic map is introduced to realize population initialization, so as to improve the quality of initial population distribution in the search space and its global search efficiency. Secondly, the adaptive T-distribution mutation mechanism is incorporated into the dung beetle position update to elevate the global exploration and local exploitation abilities of the algorithm. Finally, the equivalent partition model for active distribution network is established based on the “black box” theory to reduce the solving dimension of the algorithm and improve the location rate. The comparative experimental results of multi-verse optimization (MVO), red-tailed hawk (RTH), slime mould algorithm (SMA), and AC-DBO algorithm and the effectiveness simulation test results of hierarchical positioning model show that the average positioning accuracy of AC-DBO algorithm is increased by 15%, 10%, and 3%, respectively, compared with MVO, RTH, and SMA algorithms. After introducing the equivalent partition strategy, the average positioning time of AC-DBO hierarchical positioning model is reduced by 36.8% and 29.4%, respectively, compared with AC-DBO single-layer positioning model and DBO hierarchical positioning model and the average positioning accuracy is increased by 21% and 8%, respectively. The AC-DBO partition model has obvious advantages in solving speed, accuracy and fault tolerance, which is especially suitable for solving the fault location problem of active distribution network.