Study On MAS-Based Urban Rail Transportion Rapridship Forecast System And Application

As a way of urban public transport, urban rail transport is different to the conventional public transport in the project investment, operation features and network density. Ridership forecast is one of the important bases for urban rail transport network planning, design, construction and operation. For the shortcomings of traditional ridership forecast in the stiff human-computer interaction forms, the signal forecast model, the low computational efficiency and the intensive labor, MAS-based ridership forecast system was designed. There are four main works which listed as follow:(1) The particularities of complexity, operational characteristics and the hierarchical distribution structure of the rapidship forecast system was analyzed. According to Suzhou urban rail transport planning requirements, the scope, duration, basis and method was given, which provided the prerequisite to the design and development.(2) The architecture of ridership forecast system was designed. It contained a support layer, a application layer and a interface layer and had Workstations Agent, Line Agent and Line Network Agent. The Workstation Agent obtained the ridership forecasting information and coordinate with other workstation Agent to obtain the ridership forecasting information of the line by the prediction model. And then the Line Agent coordinated with other Line Agent to ultimately obtain the ridership forecasting information of the line network.(3) Adopted to the approach of "combination of qualitative and quantitative analysis" and "macro-control, micro-competition", from the point of traffic demand forecast and ridership forecast view, the rapridship forecast model was designed from generation model, distribution model, mode split model and traffic assignment model based on the traditional "Four-stage Method".(4) The rapridship forecast system was designed jointed with JADE platform, TransCAD, Mapinfo and Microsoft SQL Server software. The system can be effectively used to settle the complexity and distribution issues by dividing the complex rapridship forecast problem into several mutual cooperation Agent units. The traffic demand forecast analysis, network and line rapridship forecast analyses were made in close-range 2020 year. The results shown that the system has the excellence of satisfying the randomness, non-linear and non-deterministic case for the strong adaptability, robustness and flexibility.