Study On Short-time Passenger Flow Forecast And The Optimal Capacity Allocation Problem In Urban Mass Transit

With the city social and economic development and urbanization process accelerated, the contradictions generated by urban transport has become increasingly prominent, the city has restricted the major obstacles to continued economic development. It came into being when the urban mass transits, a better solution, in the city's growing traffic problem of supply and demand by the people of all ages. Therefore, the analysis of historical passenger data to predict future traffic, the passenger capacity to develop rational deployment plan for improving the capacity of urban rail transport have very important significance.This history of the Beijing urban mass transit passenger data, analyzes the characteristics of urban rail passenger traffic, passenger traffic forecasts were established and capacity allocation model and algorithm. The main work and conclusions are as follows:(1) Summary of urban mass transport network temporal and spatial distribution of passenger flow for this short time after the rail passenger forecasts and the optimal capacity allocation problem of passenger basis.(2) Use of cluster analysis and related methods such as pre-sample data to ensure that the input data and output data, reliable and reasonable, to improve the network training performance.(3) To analyze passenger flow prediction method for urban mass transit passenger flow distribution, the selected passenger flow forecast for the short program, namely neural networks, fuzzy neural network and support vector machines three methods, the use of existing data tested a short-term passenger forecasts. After a comparison it shows that the prediction accuracy of neural network maximum. So, in practice, wavelet neural network can predict more precise changes in passenger traffic, operators can make timely responses to such methods.(4) Take full account of the train lines in the limits on the amount invested, the capacity of the vehicle and passengers were stranded stations, etc., in order to put the number of decision variables subway train, the establishment of a scheduling optimization model of urban rail transit. The model through case studies, the optimal path obtained with the various cars and the company's biggest profit. The results show that a reasonable schedule, not only can shorten the waiting time of passengers, but also for reducing congestion and Metro subway operating company's costs and improve the operational efficiency of the MTR is significant.