Research On Integrated Charging Facilities And Fleet Planning For Shared Mobility-on-Demand System

The combination of electric vehicles(EVs)and shared mobility-on-demand system,which can couple the transportation and power networks,will help transportation electrification and achieve the "two-carbon target".In addition,the EV-based shared mobility-on-demand has its flexible spatial-temporal operation characteristics on the demand side.If it can be considered in the research of charging infrastructure and fleet planning,it can play the regulating role of load,realize reasonable investment allocation,and ensure the safe and economic operation of the transportation system and power system.Therefore,this paper carries out a study of integrated charging infrastructure and the shared mobility-on-demand fleet planning.The main work is as follows:Firstly,this paper proposes a joint planning scheme of charging infrastructure and shared electric fleet to optimize the allocation of charging pile sizes,distributed renewable energy resources,vehicle types,and fleet sizes at different charging rates,while accounting for the spatial-temporal demand-side flexibility of shared mobility-ondemand systems.An augmented network flow model is used to model the scheduling problem of the EV-based shared mobility-on-demand fleet,and three scheduling strategies including vehicle order serving,vehicle rebalancing,and charging schedule are optimized.Secondly,this paper proposes the joint planning scheme and establishes it as a twostage stochastic program model for the shared mobility-on-demand system taking into account the uncertainty of customer travel demands.The investment decisions and scheduling decisions of the shared mobility-on-demand system in this model are regarded as two-stage decision problems,and the optimal investment strategy is determined through the collaborative consideration of the two-stage problems.The objective function of the model is the lowest overall planning cost,including the expected value of charging infrastructure and fleet investment cost and dispatching operation cost.The model considers constraint conditions such as passenger flow balance constraint,empty vehicle flow balance constraint,travel request balance constraint,road remaining capacity constraint,charging station capacity constraint,and balance vehicle power constraint at the end time.The model not only optimizes the charging infrastructure and fleet capacity configuration but also determines the vehicle allocation and charging scheduling strategy of the shared mobility-on-demand system.Then,the Latin Hypercube Sampling(LHS)method and the multi-cut Benders decomposition method are used to solve the two-stage stochastic programming model of the shared mobility-on-demand system.The Latin hypercube sampling method transforms the two-stage stochastic model into a two-stage deterministic model.The multi-cut Benders decomposition method is used to reconstruct the model in the form of the main problem and sub-problem.The main problem of the two-stage deterministic model in this paper is mixed integer linear programming(MILP),and the sub-problem is linear programming(LP).The Gurobi commercial solver is used for the iterative solution.Finally,simulation experiments based on real-world data of Didi in Haikou City are performed,and the investment strategy,vehicle allocation strategy and charging scheduling strategy of urban planners are acquired,which demonstrates the effectiveness and feasibility of the proposed planning scheme of the shared mobility-on-demand system and multi-cut Benders decomposition algorithm.It is proved that charging infrastructure and fleet planning research can effectively maximize social welfare,and the city planners can adjust the spatial and temporal charging demand according to passenger-side travel demand,which is helpful to further reduce costs.