Hierarchical Optimization Reconstruction of Lightning Fault Microgrid Based on BP Neural Network

When the microgrid is struck by lightning, the power supply may be damaged, causing the power imbalance of the microgrid, so that the microgrid cannot operate normally. In this case, the microgrid system needs to be quickly reconfigured to remove part of the load and maintain the normal operation of the microgrid. In order to respond quickly to the emergency in MG, this paper proposes a hierarchical optimization reconstruction method of lightning fault microgrid based on back propagation (BP) neural network. This method introduces BP neural network into the hierarchical optimization algorithm, and puts the optimization process that does not involve power flow calculation in the network reconstruction process at the first level, and the optimization process that involves power flow calculation as the second level optimization processing. Only the solutions that meet the first-level optimization conditions are subjected to the second-level optimization processing, and in the second-level optimization process, the trained BP neural network is used to realize offline training, online input switching state, direct output network loss, balance node power and voltage results, and save calculation time. Finally, an improved 33-node microgrid system is built in MATLAB for simulation experiments. By comparing the algorithm in this paper with the hierarchical optimization algorithm that directly calculates the power flow, the simulation results show that the two algorithms can get the same reconstruction scheme, but the calculation time of the algorithm in this paper is very short, significantly improving the efficiency of microgrid reconstruction, thereby realizing rapid reconfiguration of microgrid under lightning strike faults.