Traffic Zone Division And OD Analysis Based On Mobile Communication Network Data

Mobile communication network (MCN) has been virtually seamless cover the major part of urban. Currently, the research on the urban traffic information acquisition methods based on mobile communication network data (MCN-D) is a hotspot in the field of traffic information detection. Taking the mobile communication network base stations (MCN-BS) as fixed traffic detectors, and mobile phones as movable detectors, they constitute a holographic traffic information acquisition system. Compared with traditional means of traffic information collection, the method based on MCN-D has many advantages, such as low-cost, low investment on equipment and high data update frequency. But the efficiency of traffic information acquisition and traffic knowledge discovery based on MCN-D is low, and the research on urban traffic land use and traffic analysis zone division (TAZD) based on MCN-D is less.Based on the above background, this paper, taking the aim of improving the efficiency of traffic information acquisition and traffic knowledge discovery based on MCN-D as the demand background, figure out the problems:how to build semantic layers to correlate the mobile communication network base station and urban transit system; how to design the method of TAZD with extracting the feature of traffic characteristics and location characteristics of MCN-BS; how to analysis the OD trips and the macroscopic situation of urban traffic system based on the above research problems.Firstly, this paper reviews the relevant information on TAZD and OD information acquisition, and summarizes the research projects domestic and overseas on extracting traffic information based on MCN-D. Through Filtering and analyzing the summaries, learn the relevant knowledge and theory, and point out the problems and shortcomings. With that, this paper detailedly introduces the mobile communication network data acquisition mechanisms and data formats. The "ping-pong switch" phenomenon of MCN-D is analyzed, and design the method of data cleaning and pretreatment.Secondly, the paper introduced the basic component of the main research contents, which is transport semantic tagging based on the MCN-BS. The traffic semantic framework is build based on the MCN. The traffic semantic of MCN-BS is tagged using the clustering analysis method, via extracting the semantic labeling features of the MCN-BS. The semantic tagging results is compared with the actual situation in Beijing to validate the effectiveness of the method.Based on these research work, this paper uses the clustering method based on weighted Mahalanobis distance dividing traffic analysis zones, through the extracted traffic characteristics features and location features of the MCN-BS. The conception of traffic zone commuting index is defined to measure the tendency of traffic zones on working or residential attribute.Finally, based on the above study results, this paper put forward the method of based on the extracted trip trajectories by means of combining Kalman filtering and time threshold to acquire traffic OD information. The validation on traffic generation, traffic attraction and traffic distribution is done in Beijing representative region. And it analyzes the crowd moving temporal and spatial situation.This research mainly study the traffic sematic framework based on the MCN-BS and tagging the traffic semantic on MCN-BS, traffic analysis zone division based on the MCN-D and traffic OD acquisition. And via the actual data of Beijing and the GIS Data Platform constructed with Supermap and Oracle, the method put forward in this paper is verified, validating the effectiveness of the method. Then it analyzes the further research questions and directions.