Research On Fault Recovery Strategy Of Distribution Network With Microgrids

As one of the important operation organizations in the future power distribution system,microgrid plays a key role in the transition stage from the traditional distribution network to the smart distribution network.With the maturity of microgrid technology and the large-scale access of distributed generation,the distribution system has gradually shown the characteristics of multi-source initiative,which provides new ideas for the restoration of the distribution network.Studying the power supply strategy of microgrid to the restoration of distribution network can avoid the influence of randomness and uncertainty of distributed generation,and effectively improve the utilization rate of distributed energy.Simultaneously,compared with traditional distribution network restoration schemes,the access of microgrids can also increase the recovery speed of the distribution network,improve the self-healing ability when a distribution network failure occurs,reduce economic losses caused by excessive power outages.Therefore,studying the failure recovery strategy of the distribution network with microgrids is of great significance to the operation and development of the distribution network.In order to be able to accurately reflect the impact of the microgrid on the restoration of the distribution network,the equivalent modeling of the microgrid is firstly studied.It is expected that the equivalent modeling of the microgrid can solve the problem of the difficulty of analyzing the operating status of the microgrid in the recovery of the distribution network,which provides a theoretical basis for the subsequent work.At the same time,the equivalent modeling method of microgrid based on Convolutional Neural Network is proposed,and the advantages and disadvantages of the two equivalent modeling methods are analyzed by comparing with the equivalent circuit modeling method,then selecting the equivalent modeling method of the microgrid in this research.Then,after comprehensively studying the development status of microgrid types,construction scale,operation control,and voltage stability,a distribution network failure recovery scheme that mainly considers micro-grid as the power supply is proposed.The following solutions are proposed for the grid-connected voltage stability verification problem of the microgrid and the distribution network load recovery strategy problem in the scheme:(1)A voltage prediction method based on the dynamic Thevenin equivalent model of the microgrid is proposed to solve the problem of transient voltage verification at the grid connection point when the microgrid is connected to the restored distribution network.This method uses the Thevenin equivalent parameter tracking algorithm based on the total differential to establish the dynamic Thevenin equivalent model of the microgrid,and then uses the BP neural network to establish the mapping relationship between the equivalent model parameters and the transient voltage of the grid connection point,realize the prediction of transient voltage by using the Thevenin equivalent model of the microgrid.(2)A method for dividing islands in distribution network based on fuzzy adaptive simulated annealing genetic algorithm(FASAGA)is proposed.A strategy has been formulated to give priority to the access of the microgrid,and then cooperate with the traditional generator sets and distributed generation to restore the load of the distribution network.This strategy uses the FASAGA algorithm to divide the distribution network into islands with maximum load recovery as the main goal.The improved FASAGA algorithm overcomes the shortcomings of the traditional genetic algorithm and greatly improves the efficiency of load recovery.Finally,the AC microgrid,DC microgrid and the US PG&E69 node distribution network are respectively used to verify the two methods proposed in the scheme,which proves the feasibility and superiority of the scheme.