Research On The Train Operation Adjustment In Urban Rail Transit Based On Fuzzy Expert System And Genetic Algorithm

Metro and light rail in the urban transit systems have unique advantages of rapidity,convenience and large capacity, which have decides that it will become one of the key pointsof unban-planning in China. Due to the impact of emergency, equipment failure or thepassenger volume, the train’s running state may become random and complicated. During thisprocess, trains need to be token effective adjustment measures and try to operation as planned,which according to the real time status’ change. There are many uncertainty factors which aresignificantly affected the train in the actual urban rail transit operation environment. However,the optimization theory is only for consideration of certainty variable. Based on these analyses,uncertainty factors are introduced into the model and algorithm to solve the problems whichare needed to be explored.Firstly, the research status on train operation adjustment at domestic and abroad isintroduced. Combined with the actual situation of urban rail transit, it analyzes the basictheories of the reasons, delay, adjustment measures about train operation adjustment. On thebasis of classical train operation adjustment model, the probabilistic model for the trainoperation adjustment with parameter uncertainties by a variety of random disturbance isconsidered. During this process, it has established the membership function for the fuzzyconstraints with the section running time, the station stop time and the track headway to traininterval operation time, stopping time and tracking interval as a core.Secondly, a fuzzy expert system is established by the basic definition and relevantfuzzy theories of train operation adjustment. According to the actual operating condition ofthe train, the fuzzy membership function of the input and output variables and fuzzy logicrules are determined. The output parameters field is solved by genetic algorithm.Finally, taking a section of Guangzhou metro line5as the simulation example, the fuzzyexpert system deals with the simulation based on the Fuzzy Logic tools supplied by Matlab,and the emulation platform which is developed by VC++6.0can realize the simulation testthat the train operation adjustment under the circumstance of departure delay. Compared withthe simulation experiment, it is indicated the validity of the fuzzy expert system, which canadjust the train operation effectively based on real-time data better. The train operationadjustment problem in fuzzy environment is explained theoretically and explored practically.Because of the influence factors, the train operation is very complicated, it needs to be furtherimproved to provide support in decision-making for train operation adjustment.