Research On Energy-saving Vrpspd With Stochastic Demands

The logistics distribution problem has extensive practical foundation as well as economical application value,and has been one of the research hotspots among the academic circle in decades.Among the numerous branches,vehicle routing problem with simultaneous pick-up and delivery(VRPSPD)has gained a wider range of concern.In reality,the level of uncertainty is largely increased because of the diversity,variability and randomness of customers.Meanwhile,the problem of environment pollution caused by delivery activities has become sharper.Therefore,how to decrease the cost of transportation,save energy and reduce pollution has become an urgent requirement to promote the delivery efficiency.So,it is of great value to study energy-saving VRPSPD with stochastic demands in both theory and practical meanings.The research focuses on energy-saving VRPSPD with soft time windows,in which pick-up demands obey Poisson distribution.The aim of the study is to increase profits and decrease the fuel consumption during the delivery,and promote the logistics enterprise's profile.Two optimization problems will be discussed in a distribution network with a distribution center,multiple vehicles,multiple customers,and stochastic demands.One is energy-saving VRPSPD with stochastic demands without considering the customers' satisfaction.Model ? is built in the goal of minimizing the difference value between cost and profit.The other one is energy-saving VRPSPD with stochastic demands considering the customers' satisfaction.On the base of Model ?,Model ? is presented in the goal of minimizing the difference between cost and profile,and meanwhile,maximizing the customers' satisfaction.For the solving of two models,firstly,this paper designed particle swarm optimization based on greedy algorithm to obtain the routing decisions in model ?.Greedy algorithm is applied to adjust the scheme to satisfy the constraints of the maximum vehicle load.Multi-Objective PSO(MOPSO)based on Pareto optimality is put forward to solve Model ?,and the result is a series of non-dominated solutions.Simulation experiment is conducted to test the effect and efficiency of the algorithm.The result of Model I shows when taking pick-up income and fuel consumption into consideration,the solutions will be different and the relation between cost and profit is positive correlation.The result of Model ? shows a negative correlation between customers' satisfaction and profit when considering service level.Paying too much attention to economic benefits can reduce the satisfaction of timeliness,and a high level of customers' satisfaction usually needs the compromise of collection income.Such conclusion conforms to the antinomy of the service level and cost of logistics distribution.In addition,taking differentiated service strategies for different customers helps the enterprise to improve the resource utilization and the management pertinence,hence establish a competitive customer relationship.