Research On Multi-Depot Joint Distribution Vehicle Routing Optimization Under Time-Dependent Networks

Vehicle routing problem(VRP)has been an issue of great interest in a wide range of fields,including transportation,logistics,and computer science and simulation technology.The focus of the VRP is to investigate the effective visiting sequence of vehicles,realize the optimal configuration between vehicles and customers,and minimize the total distribution cost under the premise of meeting a series of service constraints from customers.VRP has an extensive application background in parcel receiving and sending in the express industry,product receiving and dispatching in the commercial field,and material dispatching and transportation.Driven by the development of social economy and technological progress,China’s logistics industry has entered a period of sustained and rapid growth,and the scale of logistics enterprises has been expanding.While seeing the rapid development of logistics enterprises,we should also see the intensification of competition among logistics enterprises.Given the new environment and new problems they face,logistics enterprises must formulate distribution plans scientifically to constantly reduce costs,improve distribution efficiency,and gain an advantageous position in the competition.For this reason,this thesis focuses on four problems in the research field of multi-depot joint distribution under time-dependent networks,which are prevalent in large logistics enterprises and need to be solved,namely single type vehicle routing problem,heterogeneous vehicle routing problem,mixed fleet vehicle routing problem and refrigerated vehicle routing problem,and obtains the following results:(1)The first aim of the study is to tackle the multi-depot joint distribution single-type vehicle routing problem under time-dependent networks.Considering the limitation of vehicle speed on different types of roads in the distribution network,depot resource sharing,and the impact of real-time vehicle load on fuel consumption,an optimization model is established to minimize vehicle fixed cost,fuel consumption cost,and time window penalty cost.The adaptive genetic algorithm is designed to solve it.It firstly uses the nearest neighbor insertion method and the logistic mapping equation to generate the initial solution.Then,crossover and mutation operations are designed to improve the initial solution,and adaptive crossover and mutation probabilities are designed to improve the performance of the algorithm.Finally,CPLEX is used to verify the correctness of the model,and several groups of experiments with different scales are carried out to analyze the effectiveness of the algorithm.(2)The second aim of the study is to tackle the multi-depot joint distribution heterogeneous vehicle routing problem under time-dependent networks.Considering the influence of multi-depot and heterogeneous vehicle joint distribution,an optimization model is established to minimize vehicle fixed cost,fuel consumption cost,and time window penalty cost.The hybrid genetic algorithm with variable neighborhood search considering the temporal-spatial distance is designed to solve it.Customers are firstly clustered according to their temporal-spatial distance so that the initial solution is constructed.Then,the evolutionary operation and the variable neighborhood operation are designed to improve the searching ability of the algorithm,and the Metropolis acceptance criterion of simulated annealing is adopted to prevent the algorithm from falling into the local optimum.Finally,CPLEX is used to verify the correctness of the model,and several groups of experiments with different scales are carried out to analyze the effectiveness of the algorithm.(3)The third aim of the study is to tackle the multi-depot joint distribution mixed fleet vehicle routing problem under time-dependent networks.Considering the influence of multidepot joint distribution,mixed distribution of fuel vehicles and electric vehicles,and continuous variation of vehicle speed,an optimization model is established to minimize the vehicle fixed cost,fuel consumption cost,electricity consumption cost,and time window penalty cost.The hybrid heuristic algorithm is designed to solve it.K-means clustering method is firstly used to generate the initial population.Then the variable neighborhood search algorithm is used for local search optimization to improve the performance of the algorithm.Finally,CPLEX is used to verify the correctness of the model,and several groups of experiments with different scales are carried out to analyze the effectiveness of the algorithm.(4)The fourth aim of the study is to tackle the multi-depot joint distribution multicompartment refrigerated vehicle routing problem under time-dependent networks.Considering the influence of multi-graph between two nodes of the distribution networks,real-time traffic information on vehicle route selection,and the cost consumption of refrigerated vehicles in the loading and unloading process and transportation process,the idea of first pre-optimization and then a real-time adjustment is adopted.A two-stage optimization model is established to minimize the vehicle fixed cost,fuel consumption cost,time window penalty cost,refrigeration cost and cargo damage cost.The initial distribution scheme is obtained in the pre-optimization stage by the hybrid variable neighborhood chaotic genetic algorithm based on the historical traffic information.In the real-time adjustment stage,according to the real-time traffic information of different graphs,the strategy of updating graph selection at nodes is proposed to adjust the vehicle routes in realtime.Finally,CPLEX is used to verify the correctness of the model,and several groups of experiments with different scales are carried out to analyze the effectiveness of the algorithm.The research results of this thesis not only enrich and expand the new model,new strategy,and new solution technology of multi-depot joint distribution vehicle routing optimization research under time-dependent networks but also provide new ideas,new perspectives,and new reference for the distribution scheme formulation of logistics enterprises.