Design And Verification Of The Optimal Model For The Departure Interval Of Rail Transit Vehicles

With the rapid development of urban rail transit system,it is greatly convenient for residents to travel,and attracts large-scale passengers to enhances the attraction of public transportation at the same time.As the use of rail transit has been more and more common,the insufficient of transport capacity in rush hour and the waste in peak hour,and unexpected situation can be a problem that affect the development of urban rail.If the rail transit capacity can't be adjusted according to the real-time passenger flow,it will cause the loss of attraction of urban rail transit and the waste of resources,which violated the design that urban rail transit is used to relieve the ground traffic congestion.It has great theoretical significance to study the optimal method for the distribution of the existing lines in the big data environment,and how to deal with the emergency situation in a timely manner.Firstly,the paper describes the research background and significance of the research briefly,and put forward the necessity of optimizing the existing lines.Then,study the travel behavior of passengers,and analysis the influence factors of riding behavior on the basis of the generalized traffic behavior theory,next,used the neural network model to forecast the potential behavior when passengers take rail transit.Based on the analysis of the spatial and temporal distribution of the existing lines,a multi-objective programming model is set up with the aims of vehicle load and passenger comfort level,which helps to arrange the urban rail transport capacity reasonably.The paper enumerated big data technology,collection and prediction method of passenger flow and neural network algorithm,etc.On the basis of these theory,put forward to use the neural network algorithm for each passenger to forecast the potential travel behavior,and make statistical analysis to these behavior.Based on the prediction of passenger flow volume,the paper optimized the departure interval.Aiming at the defect of neural network algorithm that network structure is uncertain,genetic algorithm is used to optimizing.Combined with the distributed processing architecture of Map Reduce,based on the research of the parallel strategy,adopt the parallel processing in order to improve the operation efficiency,save calculation time,and prove a solution that model takes a long time in training of the optimization algorithm under the big data environment.At last,the paper made an experiment with the examples of Nanjing metro line 1,contrast the calculation time of single machine type wi th the parallel computing time,verified that the training speed can be improved by introducing parallel framework Map Reduce in the model.Calculating departure interval in normal station,and comparing designed capacity with the actual volume,to verify the rationality of the model.Simulating the bump up of traffic to verify the capacity of the model to reply the emergency.The model proved in this paper that optimize the departure interval based on the big data can adjust departure interval in real time,and provide theoretical basis for the reasonable arrangement of rail transit.