Research On Improvement Of Multi-Objective Optimization Genetic Algorithm And Its Application In Orderly Charging

In order to achieve energy saving and emission reduction,the state has vigorously promoted new energy vehicles through subsidies for vehicle purchases in recent years.As the most important branch of new energy vehicles,electric vehicles have developed rapidly,and the number of electric vehicles has increased rapidly.And large-scale electric vehicle charging will put tremendous pressure on the power system.For example,it will increase the peak valley difference of grid load,the rated load of the transformer will be exceeded.In extreme cases,the phenomenon of"large-scale electricity consumption" may occur,it will cause damage to the power grid and affect the normal power consumption of residents and businesses.On the basis of not modifying the existing power grid as much as possible,how to deal with the impact of large-scale electric vehicle charging on the power grid is a problem that needs to be solved urgently.In this thesis,the orderly charging scheduling method was selected,with the main goals of reducing the peak-valley difference of the grid load curve and reducing the user's charging cost,and the improved multi-objective optimization genetic algorithm was used to solve the orderly charging model.In this thesis,the issue of orderly charging was discussed in depth,and the main research work is as follows.First of all,this thesis constructed a disordered charging model for electric vehicles based on existing charging data sets.According to the existing data set,it was assumed that the data of all dimensions conform to the normal distribution.The model parameters were solved by maximum likelihood estimation,and the disordered charging model of electric vehicle were constructed.This thesis uses the Monte Carlo algorithm,adopted the sampling method of accept-reject sampling,and wrote a computer program to simulate the disorderly charging of electric vehicles.Secondly,this thesis constructed a mathematical model of ordered charging and improved the NSGA-?(Non-dominated Sorting Genetic Algorithm-?)algorithm.This thesis constructed the objective functions of the grid layer and the user layer,and looked for the constraints of the grid layer and the user layer.This thesis has made three improvements to the NSGA-? algorithm,namely:comparison of similar non-dominated solutions with different levels of crowding,normal distribution adaptive crossover operator(NSGA-?-ndaSBX),fuzzy clustering of large-scale Pareto optimal solution set(NSGA-?-POFC).This thesis used the improved NSGA-?algorithm to solve the ordered charging model.This thesis evaluated the improved NSGA-? algorithm,the standard NSGA-? algorithm and the MOPSO algorithm through the ZDT series of functions,and performed ablation experiments on the improved NSGA-? algorithm.Finally,experimental simulation.This thesis compared the simulation results of disordered charging,basic NSGA-? algorithm,MOPSO algorithm and improved NSGA-? algorithm through experimental simulation.The experimental simulation was proved that the improved NSGA-? algorithm was used for ordered charging scheduling,the peak-valley difference was the smallest,the variance was the smallest,and the total charging cost was the lowest,which had a better scheduling effect.