| With the continuous acceleration of the current urbanization process,people pay more and more attention to the innovation of logistics mode and the exploration of new energy.Under the background of big data and the Internet era,"sharing" and "electrification",with their outstanding advantages of energy saving and environmental protection,have become the focus of research,widely used in different fields.Especially in the urban transportation system,a faster and more convenient transportation,shared electric vehicles,is gradually replacing traditional fuel vehicles.They can not only reduce the number of private cars and ease traffic congestion,but also effectively promote emission reduction.Generally speaking,the stationbased car-sharing service is divided into two types:two-way and one-way.Among them,oneway shared electric vehicle service always causes the uneven distribution of vehicles at the station,which requires operators to send dispatching staff to relocate vehicles.Unlike the relocation of shared bikes,this job needs workers to drive redundant electric vehicles to the demand point one-on-one,rather than directly using trucks.Therefore,the repositioning of shared electric vehicles and the cooperative scheduling of workers is an important problem.This thesis studies a joint optimization of Electric Vehicle repositioning and Staff Relocation with Shuttle pick-up and delivery Service(EVSR-SS)in car-sharing.On the basis of the existing research,the staff transportation process by a shuttle is added,comprehensively considering the supply and demand matching of stations,staff tasks allocation and shuttle path optimization.The balance of staff workload is also taken into account besides minimizing the total cost.In this thesis,we propose a three-phase optimization method for this problem:(1)determining the vehicle relocation tasks under the goal of minimizing routing cost,that is,the moving path of each electric vehicle to be relocated.(2)ascertaining the minimum number of workers to be hired for repositioning and their task assignment.A bi-objective mixed integer linear programming model is established to minimize the total cost and the maximum path.(3)developing shuttle routes for picking up and dropping off workers.Moreover,a Hybrid Particle Swarm Optimization with Genetic Algorithm(HPSO-GA)is introduced,combining the ε-constraint method to solve the bi-objective problem.This algorithm combines the advantages of Genetic Algorithm(GA)and Particle Swarm Optimization(PSO),which adds the crossover and mutation process of GA to PSO.The highlight of HPSO-GA is that,it selects the parent generation with the probability related to inertia coefficient and acceleration coefficient respectively,giving the improvement direction to the crossover process to accelerate the convergence.The performance is compared with genetic algorithm,simulated annealing,ant colony algorithm and a hybrid genetic algorithm with tabu search.The results show that HPSO-GA has high efficiency and good convergence effect,and the satisfactory solution with high quality can be obtained in a reasonable solution time.In summary,this thesis can provide decision reference for vehicle relocation and personnel arrangement collaborative scheduling in electric car-sharing operation. |
