Research On Travel Behavior Prediction And Aggregator Optimal Scheduling Of Electric Vehicles

One of the keys to achieving energy conservation and emission reduction,practicing China’s sustainable development strategy and achieving the "double carbon" goal is the wide area access of large-scale electric vehicles.The charging behavior of large-scale electric vehicles(EV)brings great challenges to the operation optimization of power system.In this context,scientific and reasonable prediction of EV load,and then optimize the scheduling of EV timing charging behavior,has important practical significance for promoting the development and upgrading of EV industry and realizing the friendly interaction between EV and power grid.This paper focuses on the disorderly charging of EVs and its impact on the distribution system,the prediction of EV travel behavior based on new rough artificial neural network and the optimal scheduling of EV aggregators considering user preferences.The specific research contents are as follows:1)The influencing factors and distribution of EV charging load are analyzed,and the disordered charging model of EV is constructed.Taking the most common radial distribution network as an example,the Monte Carlo(MC)method is used for simulation solution,and the forward/backward flow method is used to realize the power flow calculation of distribution network with EV,The results show that the charging load of EVs will aggravate the peak valley difference of the system and increase the network loss of the distribution network.2)A travel behavior prediction method of EV based on new rough artificial neural network is proposed.Considering the correlation of different travel influencing factors of EVs,a feedforward and recursive artificial neural network based on artificial intelligence and a network training method based on rough structure are used to predict the travel behavior and charging demand of EVs.Use the historical data to extract the arrival time and departure time of EVs,and then predict the driving mileage of EVs to obtain the accurate charging demand of EVs.The predicted EV travel behavior is compared with the MC simulation results.The results show that the method proposed in this chapter can improve the accuracy of EV charging load prediction and is close to the real load trend.3)An EV aggregator optimal scheduling model considering user preference is proposed.Based on the prediction of EV travel behavior,an EV aggregator is introduced to optimize the scheduling of EV charging.Considering the user’s preference,three different charging packages are proposed for users to choose;the interaction between EV aggregators,distribution system operators(DSO)and transmission system operators(TSO)is proposed.Finally,taking a city’s EV charging as a case study,the advantages of coordinated charging method are quantified.The cost-effectiveness compared with uncoordinated charging is given.The results show that the method proposed in this chapter can significantly improve the income of EV aggregators,and the improvement effect is more obvious with the increase of EV penetration.