The Research Of Optimal Charging Strategy Of Electric Vehicle

With the depletion of fossil fuels in the near future,Electric Vehicle(EV)will become a potential driving mode after combined with renewable energy to reverse the situation.But,at the same time,the effective management of electric vehicles is also facing challenges.Governments,automotive electronics manufacturers,transportation management systems,and electric vehicle drivers will be involved directly or indirectly.This paper modelled and predicted the travel situation of the EV end user's,and then the result of the prediction is introduced to the battery state estimation.Analyzed the influence of travel time and travel frequency on battery capacity at different times,then introduced the SOC analysis of battery state estimation to the EV charging optimization strategy.Analyzed the influence of the power load from two angle of the terminal users travel rules and battery state estimator.The design of a smart grid charging line,the probability modeling of single EV any time in a day commuting is used,time capture different,and constructed in the time dimension of the relationship between the variables you state transtion-probablity matrix describe the driving mode and time-of-use price behavior.According to the use of the vehicle,risk aversion and end user electricity price,the sample using stochastic dynamic programming in only charging scheme and V-2-G scheme to determine the optimal charging strategy.Specific studies as shown followed:Nonhomogeneous Markov decision model(HMM)was proposed to deal with the transition probability estimation based on the practical application.HMM was used in the Veterbi algorithm combined with theoretical analysis,applied to the analysis of the driving behavior of the end users of electric vehicles,and the model is established according to the actual travel rules.To pattern recognition and weighted predict the built model.The model can be used to simulate the relationship between users travel time-travel mileage-travel frequency of the three good in the future practical application,using the maximum likelihood estimation of the similarity and the target sequence model.The main purpose is to provide a battery electric vehicle electric drive vehicles,rather than from the power grid energy storage battery.It is required to store enough energy to ensure the completion of required travel.Research on electric vehicle charging characteristics is also the study of battery charge and discharge process.According to inherent characteristics,on the basis of the use of LS-SVM on the electric vehicle charging and discharging electric charge state estimation simulation,convex optimization model Lagrange function to join my space construction is minimized,and the simulation analysis problem of battery capacity.The nonlinear characteristics of the problem itself by combination kernel function estimation.Several kinds of kernel function do effect comparison,PK-EK combination kernel function is compared with actual process.Power grid operators to develop time-of-use price is the main purpose of peak stagger,reduce network load pressure,rather than from the terminal users get higher interest.So the price divided to stimulate users to avoid the daily peak load power and low load period charge.According to design a set of intelligent power grid charging optimization route established information.The establishment of mathematical model,the decision objectives and constraints are unified,using stochastic dynamic programming from the "charging" and "V-2-G" to two schemes to determine the optimal electric vehicle charging strategy,however,were optimized.