Study on Regional Traffic Sub-Area Division Based on Improved Louvain Algorithm and Correlation Degree

Abstract

Aiming at the sub-area division problem of complex traffic network, with the goal of optimizing coordination and control, this paper proposes a sub-area dynamic division method of an improved Louvain algorithm based on an improved correlation degree model. Firstly, by introducing the traffic flow dispersion coefficient and the periodic coefficient of adjacent intersections, optimizes the traditional Whitson model. Comprehensively considering the influence factors such as road properties and saturated traffic flow density, this paper proposes a model of improved correlation degree. Secondly, uses the size of the module degree as the criterion, this paper improves the Louvain algorithm, make it is able to dynamically divide the sub-areas of the regional road network in different time periods more effectively. This study selects an actual regional road network for the simulation experiment, the results of the comparison experiment and the verification experiment are show that the proposed improved division method can better combine the traffic flow characteristics, and realize a more reasonable sub-area dynamic division for the regional road network.