Dynamic Reconfiguration Strategy Based on Partition of Time Intervals with Improved Fuzzy C-means Clustering

With the widespread access of renewable energy to the distribution network and the dynamic changes of DG and load, the dynamic reconfiguration scheme based on improved fuzzy mean clustering is proposed. First, the equivalent load forecasting curve is established based on the time variability of DG and load. The improved fuzzy mean clustering algorithm is applied for partition of time intervals, and the value of combined loss function is used to determine the optimal partition number and plan, which does not require the preset partition number or comparison thresholds. The dynamic reconfiguration model based on minimum network operating cost is established, which is solved by the decimal particle swarm optimization (PSO) algorithm based on loop search. Finally, it is proved by the IEEE33 distribution test system that the proposed strategy can effectively reduce the number of reconfiguration and network operation cost. And the algorithm is better than the traditional algorithm.