Traffic Forecasting And Scheduling Of The Urban Passenger Rail Network Based On Historical Data

Urban rail transit has become an important part of modern urban transportation system. With the accelerating speed of urban rail transit network construction, and the increasing passenger flow, the real-time features of the passenger flow has become more increasingly obvious. Therefore, it is essential to distribute and predict passenger flow with the use of reasonable assignment model and forecasting model by analysising historical passenger data. More importantly, the forcasted passenger flow can be used to compile train diagram and train schedule plans, which is of great significance to develop a reasonable scheduling operation plan and improve service quality.In this paper, taking the actual passenger data which is collected in urban rail transit automatic fare collection system as sample data, and refering to passenger flow distribution theory, passenger flow forecasting theory and scheduling optimization theory in urban public transport system, the above theory combined with the characteristics of rail transit system is applied to Beijing urban rail transit system. The main work is as follows.1. Passenger flow distribution characteristics are analysised, and a reasonable Beijing urban rail traffic passenger assignment model is established based on the actual passenger characteristics. And cross-section passenger flow of the Beijing Metro Line 1 is obtained, which can be used in passenger flow forecasting and scheduling.2. BP neural network is analysised in a great depth in connection with the unique characteristics of Beijing urban rail transit. Its practicality and rationality is researched for passenger flow forecasting. And by comparing BP neural network passenger forecasting method with the least square method used to fit the passenger flow, it proves that neural networks are better to predict the future traffic passenger in the case of a daily passenger flow distribution shows a slow growth.3. By researching and analysising train adjusting algorithm and the traffic passenger forecasting information obtained, a recursive algorithm which is suitable for the feature of Beijing urban rail transit is found. And take Beijing Metro Line 1 as an example, the daily train departure intervals are obtained through the use of this algorithm. Additionally, the average vehicle full-load ratio has been used to verify the rationality and practicality of the train departure intervals.4. Through the connectivity between MATLAB and Database, train departure times are obtained with train departure intervals. And by using MATLAB and VC Language to program train diagram simulation program, train running schedule is simulated and a complete daily train diagram is obtained.5. A graphical user interface design is used to achieve visualization of train dispatching system. Through the use of graphical user interface GUI provided by MATLAB, a reasonable Beijing urban rail transit schedule system interface is created, including the main interface, the train driver, line, station and other basic information interface and the train schedule interface. The train schedule interface contains train departure information, traffic information and parking information, and etc. All the train schedule information can be inquired and displayed by the interfaces.