Abstract:

As the urban traffic flow is a complex, changeable, nonlinear, unstructured and random system with large scale and temporal-spatial variations, the common methods with fixed threshold cannot effectively estimate the traffic movement condition. With the continuous development of the Intelligent Traffic System, it is imperative to find an estimation model which is suitable for the mixed traffic condition in China and accordant with the movement mechanism of traffic flow. In this paper, the multi-dimension state characteristics of mixed traffic flow are analyzed, and a four-dimension state estimation model is established based on the rough set theory. A two-dimension decision table is then obtained via data discretization and attribute reduction, and the rules of dimension-reduced state estimation for urban traffic system are presented by visualizing the model in a form combining both figures and tables. The results of a case study show that the proposed method can effectively eliminate the redundancy information of the system, reduce the system complexity and improve the precision of the mining rule.

Key words: urban traffic system, traffic flow, rough set, attribute reduction, estimation rule