Research On Urban Rail Passenger Flow Prediction Technology

In recent years, China's economy has been sustained and healthy development, and people's living standards has been continuously improving. Because of China's big population base, with the urban population growth and continuous expansion of city size, urban rapid growth of vehicle ownership, the layout changing of urban land use, urban traffic pressure on the increasing urban traffic congestion has become a focus of concern. To construct and develop a fast, safe and efficient urban rail transit, the key to this question is to establish the city's comprehensive transportation network system, has become an inevitable trend of the urban transportation.Rail trans has the following characteristic: Great transportation amount, quick speed, safe and comfortable, small pollutes, small area, high combined benefits. Rail transit construction is a long-term, ambitious project required immense human and material resources and financial resources to support consumption, and have far-reaching implications to the city's spatial structure and land resources. Rail transit passenger demand forecasts and project analysis is an important basis for the feasibility analysis as well as the entire rail transportation system planning and design. We must pay attention to demand forecasting methods in order to provide scientific support to the rail line network and improving.In this paper, the author summarized the development status of urban rail transit domestic and abroad and the current commonly used methods of urban rail passenger flow prediction, analyzed the existing methods of prediction. It focuses on the current use of the common four-stage transport planning method, gives improving suggestions to the use of the method of passenger flow prediction, and establishes improved prediction models and new methods of passenger flow prediction. Trip generation takes the distance into account, adds accessibility variable into the regression model. Trip distribution forecast uses modified correlation method, taking into account the actual distribution of the trip with a very strong direction, the large volume of traffic attracted to the area often comes from a relatively small number of individual traffic zones. Mode split forecast uses the four-way division technology based on competitive and cooperative relations, after analyzing competitive and cooperative relations of rail transit traffic way and other ways. In this paper, the author verified the improved methods in light of a city's actual, compared the prediction results with the prediction results of the traditional four-stage method, and confirmed the advancement and rationality of improved methods.