Time-Dependent Green Location-Routing Problem And Algorithm Research

In order to save costs and reduce fuel consumption,it is necessary to optimize various logistics activities.The Location-Routing Problem(LRP)is an important branch of logistics optimization.The classical LRP are often established in highly generalized experimental network environments.The travel time between nodes and the speed of the vehicle are constant,and the delivery distances are often used as the optimization objective.But the travel time and vehicle speed in the actual road network are time-dependent and stochastic.Therefore,the time-dependent location-routing problem has more practical guiding significance for actual logistics activities.Based on the classical LRP problem,this paper studies the green LRP problem in static network,time-dependent network,and stochastic time-dependent road network environment.The research contents are as follows:(1)The Green Location-Routing Problem and AlgorithmIn static network,the customers and the depots are nodes on the network.The distance between the two nodes is the Euclidean distance.And the speed of the vehicle is considered to be constant.Based on the static network assumption,a Green Location-Routing Problem(GLRP)was established.The optimization objective of this model is divided into three parts including vehicle costs,depots costs,and fuel consumption costs.Because the speed of the vehicle in the static network is considered to be 1,the fuel consumption is the function of vehicle load and the distance.Based on the theory of biological heuristic algorithm,an improved Artificial Bee Colony algorithm(MABC)was designed to solve the GLRP.MABC has been compared with three classic LRP algorithms.(2)The Time-dependent GLRP and AlgorithmVehicle speed has a great influence on fuel consumption and emissions.The time-dependent characteristics of the road network determine that the vehicle speed is time-dependent function.Therefore,reasonable route planning can save costs and reduce emissions.Chapter 3 of this paper fully considers the influence of speed on fuel consumption.The travel speed of each road segment uses a stepped discrete speed function,that is,different road types at different time periods corresponding to different speeds.This chapter provided a Time-Dependent Location-Route Problem model(TDGLRP)based on the classic fuel consumption calculation model CMEM.Unlike a static network,fuel consumption is dynamic in different periods.Based on the theory of biological community evolution,a Multi-Colony Artificial Bee Colony algorithm(MCABC)was designed.The MABC and MCABC algorithms were compared and analyzed based on the benchmark examples.The influences of customer distribution and time windows are analyzed.It also gives several large-scale cases with 1000 customer nodes,providing comparative data for other researchers.(3)The Stochastic Time-dependent GLRP Problem and AlgorithmThe road network in the real environment is dynamic and stochastic.The vehicle speed of each road section is uncertain and fluctuates.Therefore,in Chapter 4,the Stochastic Time-Dependent Green Location-Routing Problem(STDGLRP)model under the environment of stochastic time-dependent road network was established.In order to fully meet the requirements of the hard time windows,the stochastic time-dependent network was robust into a deterministic time-dependent network according to the maximum and minimum optimization model.In stochastic time-dependent road networks,the shortest route between two nodes is not unique.Moreover,the travel time between the two nodes varies with the departure time of the vehicle.Therefore,in a stochastic time-dependent road network environment,the shortest routes need to be calculated in real time.With different high-level strategies and dynamic acceptance mechanisms,9 Hyper Heuristics(HH)were designed.A comparative analysis of 9 combination algorithms were conducted.The number of vehicles enabled under the STDGLTP model and the TDGLRP model were compared and analyzed.And the influences of the time windows and the vehicle capacity were discussed too.As the same time the differences of the routes with different speed fluctuation functions were analyzed.(4)Multi-objective TDGLRP and AlgorithmIn the time-dependent road network,a multi-objective Time-Dependent Green Location-Routing Problem with total cost,travel time and vehicles as the optimization objective was established.And ABC Hyper-Heuristic(ABCHH)algorithm were used to solve the proposed model.