Service Restoration Of Active Distribution Network Considering Battery Charging And Swapping Station And Controllable Load

With various types of energy into the grid and load management more flexible, the form of recovery has been gradually changed from the centralized recovery to the partition recovery. Therefore, it is significant to research how to improve the efficiency of partition restoration and maximize the value of controllable loads. In the trend of clean energy development, electric vehicle will be widely used. The power storage of electric vehicle battery can provide short-term supply to the grid by the V2 G technology. This provides a new direction for the fault recovery study. Based on the above problems, this paper has done the following work:Firstly, the graph theory is applied to simplify the distribution network, and the network topology search is summarized based on the above simplification. According to the above theory, the back/forward sweep algorithm is analyzed. Considering the uncertainty of loads in dynamic recovery, the fuzzy clustering method is used to process the daily load data to figure out the training sample set. Then, the Elman neural network method is applied to make a short-term load forecast, laying the foundation for the later study.Secondly, for service restoration in multi-time intervals, a dynamic partition coordination recovery strategy considering interruptible loads is given. Each partition is determined according to the power output satisfying the demand of confidence level and the interruptible loads. The dynamic partition avoids the impacts caused by the fixed partition and the uncertainty of DG power output. Moreover, the dynamic programming algorithm is involved to search island, which improves recovery efficiency. For the independence and coordination between partitions and the particularity of interruptible loads coordinating dynamic schemes, the multi-agent system is introduced to coordinate dynamic recovery programs based on the proposed priority and contribution indexes,playing a role of the interruptible loads.Finally, the battery charging and swapping station is involved into the dynamic service restoration. The scheduling model is established according to the daily curves ofavailable batteries and total energy stored in the station and the battery swapping needs.For the time variation of dynamic recovery and scheduling of charging/swapping station,the solution takes two steps. Based on the optimal DG islands in each period, the scheduling spare schemes are obtained by searching the recombinant network excluding the islands set. The reasonable scheduling of charging/swapping station increases the quantity of important loads and improves the recovery reliability.