Research On Public Transit Priority Management And Control Based On Real-time Information

Rapid development of urban public transportion is the most effective way to improve the efficiency of urban transportation systems and relieve urban traffic congestion. However, there exist many problems in the urban transit systems such as insufficient capacity, low punctuality and low speed, which lead to the lack of attractive and competitive of public transportation systems, which is a serious obstacle of the scientific planning and sustainable development of urban transport system. The core problem is the lack of public transport system data collection and analysis methods, as well as the lack of accurate prediction model of the bus operators and management methods. Therefore, in this paper, we proposed a new method of bus priority management and control based on real-time informationâ†'bus arrival times forecast from the passenger demand analysisâ†'bus real-time scheduling â†' bus signal priority control, which is a complete process of traffic management and control. Also we design a predictive analytical models and management strategies of urban public traffic system. The main research content includes the following main points:(1) Propose a new model to estimate and analyze the passengers getting on and off data of the buses, which is able to estimate the accurate OD demand of the urban transport network, which is able to support the real-time bus scheduling and operation;(2) Study the forecasting method of the short-time bus arrival time based on real-time traffic information, and propose a LINK-CFNN forecasting model and STOP-CFNN model, which significantly improve the prediction accuracy of the bus arrival time, so as to provide passengers with travel information service and support real-time bus scheduling and signal priority control;(3) Study the bus holding scheduling method and propose a Kalman filter prediction model framework based on support vector machine, design the co-control holding strategies to improve the utilization of transportation resources, bus punctuality rate and the level of service;(4) Propose a bus signal priority control strategy based on reinforcement learning, according to the state of the environment and choose the best strategy to get the maximum value of the cumulative return, making the bus signal priority control with a self-learning ability to reduce the delays of the public transport vehicles in the intersection and improve the punctuality rate and service level;(5) Based on widely-used traffic simulation tool Paramics, traffic simulation systems of a single intersection and a bus route are built. Practical data are inputed to simulate the real traffic environment. Simulations of co-control holding strategies and bus priority control strategies were executed which proved the vaildness and efficiency of these two strategies under practical traffic environment.