Application Of Genetic Algorithm Synthesizing Ant Colony Optimization In Intelligent Bus Scheduling

Intelligent Transport System (ITS) is the developing direction of themodern transportation system and being widely studied throughout the world.Public transportation is an important component of ITS as well as anindispensable part in people's life and work. Public transportation enterprisesare responsible for the organizational management of the city's passengers'transportation, the performance of which will directly affects the overalleffectiveness of the urban function. The central task of the publictransportation enterprises is to schedule management. Through the schedulingmanagement, the operation guidelines should be retained and businessobjectives should be attained and it will directly relate to service quality,business benefit and financial costs of the enterprises. Intelligent dispatchingsystem forpublictransport is anewandmultidisciplinaryresearchfield. It canget rid of the traditional method dependent on handiworks and experience andstrengthentheorganizationalmanagementoftransportation.Ithelpstoachievethe automatic generation and optimization of the transportation dispatch andperfectoperationservicesandmanagement.The subject that Genetic algorithm deals with is not the parameter itself,but theindividual that encodes the parametersets. Andthis algorithm operatesdirectly on the composition target. Meanwhile it also operates on manyindividuals in a group, that is, it evaluates many solutions in the search spacethus reducing the risk of getting into locally optimal solution and this geneticalgorithm is of easy implementation of parallelization. Genetic algorithmbasically does not need knowledge in search space or other auxiliaryinformation, but only to employ fitness values to evaluate individual uponwhich the genetic manipulation is carried out. The fitness function is notrestricted by the continuously differentiable and its domain can be set atrandom.Geneticalgorithmadoptstransitionrules ofprobabilityinsteadofthedeterministic rules to guide the search direction. Genetic algorithm ischaracterized with self-organization, self-adaptation and self-education. Basedon these very characteristics of the Genetic algorithm, this thesis adopts thisgeneticalgorithmasonewaytodealwithpublictransportdispatchproblem.The main characteristic of Ant colony algorithm is based upon the self-organizational collective forage behavior of real ants of which the primaryfeatureis that ants canseekout theshortest route between thefood source andtheir net through collective behavior. When the ant moves in between thefoods and the hole, it will give out some volatile pheromone, so other antssmell this enzyme and follows the right path, and they will finally convergeinto the same path and move on. Ant colony algorithm is an effective andintelligentmethoddealingwithcombinatorialoptimizationproblems.If the Ant colony algorithm and Genetic Algorithm are to be combinedandiftheAntcolonyalgorithmistobeappliedtosteerthemutationofgeneticoperator,shouldthecombinatorialoptimizationproblemsbesolvedeffectively?With this question, this article makes a study on the hot topic of intelligentmanagementdispatchsystemandtheprimarytasksaresummedupasfollows:1. Starting with ITS, this paper elaborates on the generation andcompositionof ITS,its basiccharacteristics andfunctions, thestatus quoofitsresearchanddevelopmentanditsresearchdirection.Onthebasisofthose,thispaper introduces the intelligent dispatching system for public transport andthenthestructureandtechnicalbasis.2. Through studying the basic knowledge of genetic algorithm and antcolony algorithm, the method of blending the two is discussed in this paper.Hybrid algorithm is the complement in superiority with the purpose ofachieving the win-win situation of optimal performance and the timeperformance. Through research, the Genetic algorithm is found to have theabilityof extensive and quick global research but can not make enough use ofsystemoutputinformation.Itdoesmakealargeredundantrepeatsofsearchingwhen solving is to certain scope. As a result the efficiency to obtain preciseresults is reduced. Ant colony algorithm is a mechanism of positive feedbackthat is lacking in incipient pheromone and takes longer time for solution. Bybringing the superiority of the two into full play, and after the crossoveroperationof geneticalgorithm,theoptimal chromosomes arepickedout as theexcellent solution and the ants are used to find better chromosome. Every antselects route in the same way as the selection of probability function andoriginates a new chromosome. If the new is of better adaptability than theformerly excellent one, it is kept and treated with Cauchy mutation operationintoanewone;ifnot,itwillnotbekeptandoneantislefttoremainsearching.The search stops only after the number of the sought excellent chromosomereaches the population size. In this way, the ant colonyalgorithm is embeddedinto the Cauchy mutation operation of the genetic algorithm, thus making theCauchymutationoperationmorepurposefulandintelligent.3. This paper discusses some practical problem in public transportdispatching system and suggests the adoption a new means of combining thegenetic algorithm and the Ant colony algorithm to deal with those problems.The intelligence of Genetic algorithm and Ant colony algorithm is fullyemployed to arrange the scheme of public transport dispatching and Designandrealizethehybrid algorithm directedat each linkofpublictrafficschedule.Genetic Algorithm (GA) is a global optimal searching algorithm suitable forthe solution of the massive, discrete and nonlinear problems in vehiclescheduling and it is of fine global optimization and robustness so that it canobtainquicklyandeffectivelytheoptimalsolutioninvehiclescheduling.Astothe Cauchy mutation operation in genetic algorithm, the basic Cauchymutation operation is arbitrary which combines the genetic algorithm and antcolony algorithm. Upon those basics, the ant colony algorithm is used as theguidance genetic operation and to enhance the intelligence of geneticalgorithm, which is original in this thesis. The mathematical description,operating procedures and programming of intelligent dispatching system forpublic transport are presented in this thesis with one bus as a case study. Theexperimental result in the mathematical simulation indicates that it is correct,reliable and effective to adopt the hybrid algorithm which combines theGenetic algorithm and Ant colony algorithm to realize the intelligentdispatching for public transport. By this means, the desired results can beachievedsatisfactorily.4.It can be believed that to optimize public transport dispatching bymeans of the hybrid algorithm of genetic algorithm and ant colony algorithmwill bring about sound distribution of public vehicle resources and adjust thebalance between supply and demand. It will improve transportation efficiencyand provide the necessary technical support for the construction of excellentschedulingmanagementmeasuresforpublictransport.