Study On Optimization Of Service-oriented Shuttle Bus Routing Problem Under Internet Background

Shuttle bus service is one of the welfare conditions provided by enterprises to employees.Employees' travelling experience directly or indirectly affects employees'satisfaction with enterprises.How to plan the route of the shuttle bus more efficiently and practically and choose a reasonable car model plays a very important role in improving employees' travelling experience.At the same time,with the rapid development of Internet technology,the introduction of emerging technologies including the Internet of things and artificial intelligence in the shuttle bus planning is conducive to further improve the timeliness and efficiency of shuttle bus planning.Based on the employee's travel experience and the optimization goal of reducing the travel time of employees,this paper proposes the optimization of bus routes for serviee-oriented enterprises.In order to further fit this Problem,we present the Weighted Vehicle Routing Problem(WVRP)model that describes this Problem,and the model introduces Vehicle load and driving distance into the objective function.The minimization of the objective function under the WVRP model means that the average travel time of all employees is minimized.The planning of the driving path of the shuttle bus under this model is conducive to improving the travel experience and satisfaction of employees.Based on WVRP model,this paper presents an ILS algorithm to solve the model.This algorithm constructs the initial solution based on the regret value insertion method,introduces seven neighborhood search operators to expand the search range of the solution,and sets the floating range of the solution in the disturbance stage to jump out of the local optimum.In order to verify the quality and stability of the algorithm,we test the algorithm with three sets of examples from Benchmark standard example library,and compare the RI-ILS algorithm with the BEAM-MMAS algorithm for WVRP model in the existing literature.The experimental results show that the RI-ILS algorithm is superior to the BEAM-MMAS algorithm.In addition to testing the examples in the standard library,this paper also verifies the effect of the algorithm in practical problem solving by using the data of enterprise shuttle buses in Tianjin and Dalian provided by Yitong innovation Dalian technology co.,LTD.In order to meet the actual transportation situation,we set different models for solving,and gave the optimal scheduling scheme under each model.The comparison between the scheduling scheme solved by the RI-ILS algorithm and the scheduling scheme solved by the BEAM-MMAS algorithm shows that the overall cost of the scheduling scheme given by the RI-ILS algorithm is reduced by 35%on average.In order to adapt to some large enterprise existence of two or more than two factories,namely synergy to ferry staff between multiple factories,WVRP model is presented in this paper,on the basis of the extension of the single station yard,up to the first time put forward the yard to take more goods weight Vehicle path planning Problem(Multi-Depot Weighted Vehicle Routing Problem,MDWVRP),and established the mathematical model of the Problem.Meanwhile,in order to solve this model,this paper designs a simple and efficient multi-starting iterative neighborhood search algorithm(MSILS)based on neighborhood search algorithm to solve the problem.MSILS algorithm adopts an efficient neighborhood structure,which is composed of a series of basic neighborhood operators,and improves the search efficiency of the algorithm.In addition,in order to avoid the search process falling into local optimum,a perturbation mechanism of solution reconstruction is designed.In order to verify the effectiveness of the algorithm MSILS,this paper conducts a test based on 33 open standard cases provided by Benchmark library,and compares them with two methods in the existing literature(BEAM-MMAS and VNS),and analyzes the results from four dimensions of solving quality,running time,stability and difference of statistical significance.The results show that the MSILS algorithm is superior to the existing algorithms in solving quality,running time and stability.Since both MSILS and VNS are based on the neighborhood search framework,in order to verify the significant difference between the two algorithms in solving effect,we also give the Wilcoxon Rank Sum Test between the two algorithms.