Clear and effective information extracted from large structured, data-rich, weighted and directional networks is a fairly common problem.In this paper, a general and simple method proposed by Thomas Louail was used to divide the weighted OD matrix into four different parts and simplify it into a 2×2 matrix, each part representing a kind of travel flow.With the data of Beijing mobike, OD matrix from 6 am to 10 PM on weekdays was extracted to compare the changes of each type of flow over time, and then the spatial and temporal distribution of Shared bikes was analyzed to obtain the travel pattern characteristics of Shared bikes.Generally speaking, the first feature is the subway station as the center, and Shared bikes play an important role in the first/last kilometer of people's travel. The other feature is that they undertake People's Daily short-distance travel.The research results of this paper can help managers and governments better understand the travel characteristics of Shared bikes, so as to better arrange the delivery and management of Shared bikes, and the government can also plan relevant policies according to the needs. Keywords: Shared bike OD matrix travel characteristics

CICTP