Research On Microgrid Energy Management System Based On Multi-objective Optimization

With the rapid development of distributed generation technology,the application of microgrids has also received increasing attention.Distributed power source operated in the form of microgrid can effectively increase the utilization of wind power and other intermittent power,ensure the important loads can supplied uninterrupted,and provide electricity power for remote areas.The microgrid energy management technology is the key technology of running microgrid economically and stablly.Microgrids can integrate and coordinate internal resources better,make full use of various sources,and improve the stability and economy through the energy management system.Due to many optimization goals on both sides of the supply and demand when the microgrid is connected to the grid,the optimization model is more complicated in such problem,and it's difficult for the model optimization at present.The randomness of the intermittent power output can easily cause system fluctuations and power imbalances when the microgrid operated under island mode.In view of the above problems,the energy management of grid-connected and island operation modes of the microgrid has been deeply searched in the thesis.The main work is as follows:First of all,the various types of microgrid structure and control methods were discussed,then the mathematical models of objects managed by microgrid were introduced and analyzed.Considering the strong generalization ability,high efficiency and easy implementation of the random forest algorithm,the random forest algorithm was used to predict the data.Then the design and solution methods of microgrid's multi-objective optimization model based on Pareto were discussed.In order to cope with the impact of new energy generation randomness on operational stability,flexibility of the microgrid has been introduced as one of the goals of the multi-objective model.Because high correlation of the photovoltaic sunrise force curve,the concept of correlation initialization is proposed and applied to the multi-objective optimization algorithm to improve the algorithm's convergence speed optimization ability and reduce the time complexity of algorithm.In light of the Pareto theory and the NSGA-II multi-objective algorithm for the optimization of economic costs,environmental costs,and load curve during the operation of the microgrid,a multi-objective scheduling process that can optimize the Layered multi-objective optimization algorithm for aspect requirements comprehensively was proposed.The multi-layer target loop optimization process ensured the global optimization ability of the algorithm and makes the algorithm suitable for solving the problem of optimization on both sides of the supply and demand of the microgrid.The proposed ASSNSGE algorithm introduced the assimilation process to accelerate the convergence speed of the algorithm;the isolation crossover mutation improves the Pareto frontier distribution of the algorithm and improves the performance of the algorithm effectively.Finally,in light of the case of in a factory's microgrid which consisted with PV,storage and diesel engine,energy management system framework and functional modules at each management level were designed for the microgrid,and an energy management system optimization scheduling and data prediction interface was designed in Visual Studio(VC)platform.The energy management function was simulated in MATLAB.