Research On Dual-body Scheduling Optimization Of Shared Electric Vehicles Under Incentive Mechanism

In recent years,shared electric vehicle projects have been very active in the field of public transportation,and large and medium-sized cities such as Shanghai,Beijing,and Hangzhou in China have vigorously promoted them,and have achieved remarkable results.With the help of shared electric vehicles,users can enjoy the convenience of private car travel with less cost and complete point-to-point displacement.However,the user's travel is tidal,which also creates an imbalance in the spatial and temporal distribution of shared electric vehicles,resulting in the phenomenon that some stations have no car available,and some stations have no place to stop.For operators,the mismatch between the supply and demand of the site vehicles or the unreasonable inventory range will cause problems such as lower user satisfaction and lower system service levels.Based on above situation,after fully considering the influence of factors including walking distance and travel costs on the user's travel plan selection behavior,a vehicle scheduling optimization method based on reward mechanism that guides the user to cooperate with the vehicle dispatcher is designed to achieve rebalancing of the shared electric vehicle system with minimum cost.This article first analyzes the relevant factors that affect the user's travel path selection behavior,and on this basis,completes the design and arrangement of user's intention questionnaire.Through the statistics of the sample data results,a multiple logit model of user travel plan selection based on three attribute variables of travel cost,walking distance after getting off and discounts is established.Further,according to the model and the utility maximization theory,the corresponding user reward mechanism is formulated.Then,based on some assumptions,a dual-body allocation optimization model for shared electric vehicles under a reward mechanism is established.Under the premise of meeting the time window constraints,battery power constraints,and departure station constraints,according to the corresponding reward rules,guide users to adjust their original travel plan,so that they can coordinate with dispatchers to complete vehicles scheduling task,so as to achieve the goal of minimizing total dispatching cost,including reward costs,vehicle power consumption cost,and penalties for outstanding orders.Considering research problem and the characteristics of the model,this paper uses an improved genetic algorithm to solve the problem.This algorithm retains the traditional genetic algorithm solution process,and makes some adjustments to initial solution generation,chromosome coding,selection,crossover and mutation operations.Finally,the high-tech park in Dalian City,Liaoning Province was selected as the research area,and some historical order data of a car-sharing application system was collected for example analysis.The departure station in the historical user order data is same as the departure station in the dispatch order as the collection of potential user participating in vehicle scheduling.Genetic algorithm to solve.In order to verify the superiority of this strategy,this paper compares it with the operator-only and user-only vehicle scheduling methods.The calculation results show that this method can further reduce total dispatching cost and improve service efficiency under the premise of satisfying all vehicle scheduling requirements in the system.