Research On Multi-objective Programming And Decision Making Of Distributed Power Grid

The combination of distributed generation(DG)and centralized power generation is the development direction of electric power industry,and it is also an essential part of building smart grid.The access of DG converts the distribution system from the single radiation network distribution into middle or small distributed active network,which has a great influence on the node voltage,power flow,reliability and network loss of distribution network.Obviously,the influence is closely related to the location and capacity of DG,which is reasonable or not directly related to the economy,security and stability of distribution network.Hence,the paper pays attention on the studies about the optimal allocation of DG,differential evolution for multi-objective optimization(DEMO)and also grey correlation analysis are studied to provide some effective optimization method for multi-objective programming and decision making of distributed power grid.The main contents of this paper are as follows:(1)This paper summarizes the current research status of DG configuration at home and abroad,and analyzes the influence of DG on the distribution network planning and operation.Based on the basic requirements of power system operation a multi-objective optimization model is established to minimize the active power loss,voltage deviation and static voltage stability of distribution network,which has considered the economy,security and stability of distribution network operation.(2)Based on the DEMO,an improved differential evolution for multi-objective optimization is proposed.Differential evolution has weaknesses of lacking sufficient selection pressure when handling complex real-world optimization problems.Therefore,a ranking mutation operator is designed based on non-dominated sorting and crowding distance,which strengthens the selection pressure of the algorithm and improves the search success rate of the algorithm.The self-adaptive adjusting control parameters is introduced into the algorithm to avoid the tedious process of choosing suitable control parameter that improved the robustness of the algorithm.Phasing out the minimum crowding distance individual strategy is adopted to improve the diversity and uniformity of Pareto solution.The improved DEMO has been evaluated on ZDT test functions,compared with similar algorithms,the results show that the algorithm has better convergence and diversity.(3)To evaluate the rationality of the model and the proposed optimization algorithm to solve the DG optimization feasibility model,taking IEEE33 system for example,respectively by NSGA-II,DEMO and improved DEMO algorithm for solving DG grid multi-objective optimization model,and compared with DEMO and NSGA-II from the aspects of outer solutions,C metric and S metric.The results show that the optimization model can better reflect the effect of DG on grid distribution network,and the improved DEMO algorithm has better performance.(4)Grey correlation analysis is utilized to obtain final scheme from the non-dominated set.In order to make it more scientific and reasonable,the subjective and objective weights of each attribute are obtained by using triangular fuzzy analytic hierarchy process and entropy weight method and integrated by game theory that takes into account the decision makers to determine the value of each attribute and the amount of information carried by the objective data.The result shows that this method can be used to select the program which can make the decision maker satisfied,and achieve the final decision of the DG multi object location and sizing scheme.