Path Optimization Of A Parallel Mobile Charging Service For Shared Electric Vehicle Clusters

In recent years,with the rise of "Internet+" and "Sharing Economy" concepts,Shared Electric Vehicle(SEV)are being promoted and popularized in major cities across China.However,as users travel on a daily basis,there are significant tides and patterns,and clustering(i.e.,a large number of SEV parked at a particular site)in popular places such as supermarkets and shopping malls.This means that SEV are unevenly distributed,i.e.,some sites have a large number of SEV while others have no or only a few SEV,and operators need to replenish SEV in time to ensure that they can serve more customers.Therefore,how to give full play to the advantages of SEV cluster distribution and improve the charging efficiency of SEV becomes a problem for operators to consider.Based on this,this thesis introduces Multi-mode Mobile Charging Vehicle(MCV)to provide flexible charging services for SEV.Then the MCV mobile service path,charging station layout and customer parking point location are optimized respectively to help operators reduce charging cost as much as possible.The main points of this thesis are as follows.1.Service path optimization for multi-mode mobile charging vehicles.In view of the phenomenon that free-floating shared electric vehicles will have cluster distribution during use,so in order to fully utilize the advantages of small spacing and overlapping idle time windows of SEV in cluster distribution,a multi-mode mobile charging method consisting of high and low service efficiency MCV is proposed,and by analyzing the distribution location and aggregation of SEV,the charging problem of SEV is then converted into an MCV mobile path planning problem considering time windows,and then an MCSP-TW-MMS model is constructed for this problem.The model is solved to rationalize the number of mobile charging vehicles and service routes for two different service efficiency.In addition,the case of refusing to provide services to SEV when operators cannot profit from these services is simulated,and the corresponding penalty cost is set.Finally,the reliability of the model is tested through case studies.2.Parallel service order optimization for mobile charging vehicles.Given that SEV in cluster distribution are often parked centrally at the same station,their spacing is generally small and their idle time windows overlap.In turn,MCV have the ability to charge multiple SEV at the same time.Therefore,the cluster nodes are divided according to whether they can be charged by one MCV at the same time period,while taking into account the case when the MCV is low on power and needs to go to a charging station for charging.Then,parallel service path planning was performed for MCV.Based on this,a mixed integer linear model is constructed.In addition,a generalized cost function considering the time window violation penalty is constructed to explore the solution space during the solving process,and some new removal algorithms are proposed to improve the efficiency of the solution.Finally,the efficiency of the algorithm is demonstrated by a comparative analysis with the optimization solver CPLEX.3.Collaborative optimization of charging station layout and mobile charging service path.In the process of charging the SEV by MCV there is a need to go to the charging station and to save time,the MCV will only replenish the power according to the subsequent service demand.And since the corresponding mobile consumption cost and power cost will be generated in this process,it can be considered that the charging path and charging volume of MCV are very important for their service path planning.Also considering the uneven supply and demand among shared electric vehicle sites,a scheduling scheme combining integrated dispatcher scheduling and user scheduling is introduced.Then the service location-path planning model of mobile charging vehicles is constructed.And in the solution process,the improved genetic algorithm model is firstly used to solve the solution.And on the basis of all kinds of practical factors into consideration,the RIL method is used to linearize the model,thus ensuring the existence and uniqueness of the final solution.4.Collaborative optimization of customer parking points and mobile charging service paths.In order to improve the situation of uneven supply and demand of SEV and over-distributed distribution among different regions,this thesis makes users participate in vehicle scheduling by introducing a user incentive mechanism.At the same time,given that there is a mutually constraining relationship between the cost of reducing the power consumption of MCV and the input cost required to incentivize users.Therefore,a two-layer optimization model is constructed for the two-subject problem of collaborative optimization of customer parking points and mobile charging service paths,and then the upper and lower layer models are solved using the column generation algorithm and the SCE-UA algorithm.Finally,the feasibility and effectiveness of the model and algorithm are proved by case studies.