In this research, Mobile phone data and road networks topology data are used to study highway networks status, especially the detection of critical road. First to map the spatial and temporal properties of the data with a mobile phone to the road network, can be used to build a road network topology. Then use the user's displacement over a period of time to generate the OD matrix, community detection algorithm Infomap algorithm needs to divide the n nodes in the network into m clusters in order to minimize the expected description length of the random walk. This principle can detect critical links. we used one month of mobile phone data covering the road network formed by the two highways in Beijing. Experiments show that the detection of road networks data with different weights will result in different types of critical link segments. The critical link overlap ratio obtained by the road network with a weight of 1 and a weight of flow is only around 60%.

CICTP