Optimal Allocation Of Distributed Generation In Distribution Network Based On Improved NSGA2

With the continuous development of the power industry in China,the defects of the original centralized and large power grid have gradually emerged.The development of distributed generation is called the inevitable requirement of the sustainable green development of the society.All kinds of distributed generation technologies have developed rapidly.However,the integration of distributed generation into the distribution network will have a variety of effects.On the premise of controlling some adverse effects,It is an urgent problem to make the best economic and ecological effects of the distributed generation grid connected to the network.In this thesis,an improved NSGA2 algorithm is proposed to optimize distribution of distributed generation supply in distribution network.First,the improved K-means clustering algorithm is first introduced to the new parent selection process of NSGA2 algorithm,which solves the blind area of Pareto optimal solution and which is easy to cause by the original crowding mechanism.Second,in the elitist preservation process of NSGA2 algorithm,we directly eliminate the inferior non dominated layer,thus ensuring the trend of approaching the optimal solution in the process of optimization iteration.Because the initial clustering center of the K-means algorithm is generated randomly,it has great volatility.In this thesis,the TD(Tree Distribution)algorithm is used.The initial clustering center of K-means is set,and a group of representative initial cluster centers is obtained by pruning the smallest spanning tree of the individual.To solve the volatility of K-means algorithm,we can better introduce it into the NSGA2 algorithm,eliminate the blind area of Pareto optimal solution,get a set of representative and diverse Pareto solution set,and give managers with better solutions.For the optimized NSGA2 algorithm proposed in this thesis,the simulation of the solution of a double objective function is carried out.The simulation results show that the algorithm is superior to the solution of the dual objective function.The blind area of the Pareto optimal solution is effectively eliminated.Then the MATLAB software is used to configure the distributed generation supply for the IEEE33 node system using the NAGA2 and the optimized NSGA2 proposed in this thesis.Two dual objective functions are constructed to find the optimal solution for the location and capacity of distributed generation supply from the three aspects of economic cost,power quality and environmental friendliness.The simulation results show that,compared with the NSGA2,the optimized NSGA2 can get a better set of Pareto solutions.To a certain extent,the optimization blind area of the NSGA2 algorithm is eliminated,which makes the solution set more diverse.