The Research On Bus Scheduling Based On Intelligent Optimization Algorithms

With the rapid development of cities in my country today,the urban population has become denser with the rapid development of cities,and urban transportation planning has become more and more important.Vehicle scheduling is an important part of urban transportation planning.Vehicle dispatching generates a reasonable dispatching plan for public transportation vehicles,assigns a reasonable departure time to each public transportation vehicle,and constructs a task sequence of vehicles throughout the day.A good vehicle scheduling method can greatly reduce the operating costs of the transportation company and improve the service quality of the bus company.This article focuses on the topic of bus dispatching,and studies two practical problems of public transportation:bus dispatching under uncertain conditions,bus dispatching with branch lines,mainly including the following work:(1)Propose a vehicle scheduling method under uncertain conditions based on a non-dominated genetic algorithm.First,a candidate vehicle block set generation algorithm is designed to generate a vehicle block set.Secondly,a vehicle block subset selection method based on a non-dominated sorting genetic algorithm is designed.Vehicle block subsets are selected from the candidate vehicle block set,and each vehicle block subset represents a non-dominated solution.Finally,the departure time adjustment algorithm is used to improve the non-dominated solution set to further improve the quality of the solution.(2)Propose a vehicle scheduling method under uncertain conditions based on the decomposition multi-objective evolutionary algorithm.This method and the above scheduling method based on non-dominant sorting are applied to Nanjing’s bus lines,and compared with other methods.Experiments show that the scheduling schemes obtained by these two methods have a lower number of uncovered time points and better results.In the first experimental scenario,10 uncovered start times can be reduced,and in the second experimental scenario,3 uncovered start times can be reduced.(3)A vehicle scheduling method based on the ancient metallurgy technique based improved simulated annealing algorithm is proposed to solve the problem of bus scheduling with branch lines.First,a code for this problem is designed.Secondly,a method of randomly generating the initial population is designed to generate the first generation population with high diversity.Then,a population evolution method based on the Nippon Sword Annealing Algorithm is designed,and the folding operator is improved to improve the convergence speed of the algorithm for the vehicle scheduling problem.Finally,an adjustment method of departure time is used to improve the quality of the archived centralized solution.This method is applied to the actual bus lines in Qingdao.The experimental results show that compared with the traditional annealing algorithm-based vehicle scheduling method,the scheduling scheme generated by this method can reduce 3 uncovered start times and 1 recovered start time.