Optimization Research On The Train Headway Time In Urban Rail Traffic Based On Fuzzy Neural Network

Because of the technical advantages of large capacity, fast speed, low pollution,safety, on time and so on. Urban rail transit can solve the problem,which is the growingcontradiction between supply and traffic demand, much better. Urban rail transit havebecome a very important part of urban transportation system. Considering urban railtransit as a public transport, its main effect is to remission urban traffic pressure andsatisfy the demand of passenger’s daily trips. Therefore, as a very important part of theoperation of urban rail transit, train headway time have very important significance forpassengers attraction and service levels improved.This Paper considering from the passengers, and referring to operation, passengerflow forecasting method and scheduling optimization theory in urban publictransportation system. Then, combine with the characteristic of urban rail transit toresearch the train headway time optimization. And then, using the simulation softwareplatform to analysis the result.The main work is as follows.At first, based on the analysis of the passengers flow distribution characteristicsand the main quality in the urban rail transit system. And, referring to traditiondetermination method of train headway time, considering about its deficiencies. Then,to create a train headway time optimization mathematical model which is based on thepassenger satisfaction and enterprise satisfaction. And theoretical analyzed therelationship between headway time and passenger satisfaction, enterprise satisfaction.And then, using the intelligent method which is fuzzy neural network to researchthe problem of urban rail transit headway time. Referencing for the establishment ofmathematical model, building a fuzzy neural network model. After introducing thetheory about fuzzy control and neural network, analyzed and established a fuzzycontroller and neural network model which including determine membership function,select fuzzy rules, structure establish and spread of neural network. Then, established afive layers fuzzy neural network model which has two input and one output.The last, simulating by MATLAB, and analyzing the simulate results. In this partof paper, using MATLAB toolbox, and programming the fuzzy neural network model,and then, training the network by example. At last, to experiment the trained network, then using the result to prove that the fuzzy neural network method is applied in urbantransit headway time.