Optimization Of Electric Vehicle Route Selection And Charging Navigation Strategy Based On Crowd Sensing

Since entering the 21 st century,human society is faced with oil crisis,energy security,environmental protection and many other challenges.More than 100 countries,including China,adjusting the energy strategy,new energy vehicles development plan and related policy was proposed.Electric Vehicle(EV)is widely followed by using electricity,solar energy and other green energy instead of oil as power.However,since the power supply and operation characteristics of EV differs from ordinary fuel vehicles,it is difficult to reach the destination when the EV on the way for the battery capacity is not enough,then it needs to enter the fast charging stations for emergency charging.At this point,the EV user should not only consider the charging styles,charging facility’s location and its using state,but also the road traffic distribution and traffic flow information.As a result,the EV route selection and charging navigation optimization helps to improve the efficiency of the user’s travel,avoiding massive impact on the operation of the power distribution system because of EVs’ charging or discharging,is of great significance to promote the rapid popularization of EVs.Firstly,this paper introduces the application of data mining technology in the intelligent transportation system,and then presents the framework of EV route selection and charging navigation based on crowd sensing.With the aid of traffic data the EV users upload to decision center,we will get real-time traffic and charging station information.With time and sections for horizontal ordinate respectively,the road velocity matrix is constructed.For the matrix may be incomplete,all velocity matrix elements acquisition based on matrix factorization method is realized,and ensuring the accuracy of the matrix factorization.Using the matrix of road velocity’s characteristics of space and time,by means of multiple stepwise linear regression and auto regressive integrated moving average model respectively,the velocity matrix restoration for the presence of empty row or column is realized.Considering the users’ average arrival rate and the average service rate in the fast charging station,the users’ average waiting time estimation method based on queuing theory is given.Based on the standard electricity price(such as fixing price or time-of-use price belongs to power distribution system)used in fast charging station,the real-time electricity price formulation method in the fast charging stations based on crowd sensing is presented,aiming to reflect the dynamic change of EVs charging load in fast charging station.Secondly,the optimal travel time,optimal charging cost and integrated optimal strategy from the initial position to the destination are proposed in the EV route selection and charging navigation model.The path selection,time,battery capacity,charging or discharging state mutual exclusion constraints are taken into consideration.Case studies are carried out in a city center within 100 × 100 km zone where the system comprises four fast charging stations and an IEEE 33-bus power distribution system.In a single private EV user,for example,for considering and not considering the real-time traffic information and serving information in the fast charging station,the EV route selection and charging navigation optimization results are compared in different price mechanisms as well as different proportion of EVs participating in crowd sensing,analysis is taken out in different decision goals affects the travel path of the user and the EVs charging or discharging effects the power distribution system.Finally,the EVs’ driving requirement and the double specific characteristics of charging or discharging are considered,for the EVs participating in logistics distribution,for example,the optimal EV user’s travel time and fast charging cost strategy are proposed in the logistics distribution model.The route selection,time,battery capacity,cargo capacity,as well as the number of EVs participating in the logistics distribution constraints are taken into consideration.An urban area within 60 × 60 km zone contains 28 client / 4 fast charging stations / 1 distribution center of 33-bus logistics distribution system,simulations are carried out to show the effect of customers’ demand,real-time traffic information and electricity price information to EVs route selection and charging navigation in the logistics distribution and its influence to the fast charging stations,optimization results have demonstrated the feasibility and effectiveness of the proposed model.