Application Study Of Evolutionary Algorithm On Bus Dispatching

With the accelerate development of world urbanization, urban traffic problem hasincreasingly become the major problem affecting social production and people’s life. Allthe countries started to pay more attention to the operation and management oftransportation system, how to effectively solve the traffic problems becomes the focus ofnational attention. As one of the important ways to solve the problem, gradually got themore attention and research from scholars, and obtained certain achievements, this paper ison the basis of evolutionary algorithms to discuss model of bus dispatching problem.This paper briefly describes the development background of intelligent transportationsystem, then detailed introduces the bus scheduling theory research and the presentsituation. As a kind of multi-objective optimization problems, before making busdispatching, several aspects should be considered at the same time to achieve the optimalstate, this paper probes into the development history of multi-objective optimizationproblem, bus dispatching and multi-objective optimization are discussed in detail. Thenbriefly summarizes the basic content of genetic algorithm, and the solving methods ofmulti-objective optimization problem are studied through genetic algorithm. The NSGA-IIalgorithm is discussed in detail, based on fast-non-dominated-sort and crowding-distanceof NSGA-II algorithm, a new mutation operator is improved in this paper, by applying thefunction test, the results of numerical experiments show that these improvements will beable to get a more reasonable distribution solution. Driving dispatching as an importantbasis for intelligent transportation system, it has significant meaning to relax the urbantraffic problem. The paper establishes a new public traffic dispatching model, in the model,both passenger’s and bus company’s behalf are considered.At last, this paper summarizes all the work and points out the research direction in thefuture.