Special Column on New Theories, Methods and Practices of Road Traffic Control
DING Fei, LI Xiang-yuan, LYU Yan, WANG Ye, JIANG Lin-yuan, JI Hui, TONG En, ZHANG Deng-yin
With the continuous acceleration of urbanization, urban functional zoning has shifted the emphasis from scale growth to quality improvement. Optimizing urban spatial layout, deepening multi-mode integration of traffic operation, and building an integrated travel service platform are the core needs of urban transportation's digital transformation. Urban traffic travel characteristic mining and behavior analysis are helpful in improving the urban multidimensional transportation service system, meeting diversified travel needs, promoting the rational development and utilization of the urban land, and guiding the urban decision makers to formulate reasonable planning measures. Cellular signaling data (CSD) has the advantages of wide coverage, large sample size, and long-term continuous monitoring. Cellular-network big data can analyze the origin-destination (OD) distribution and travel behavior pattern of individuals or large populations at a lower cost, thus being important for promoting the development of future urban intelligent transportation. In this paper, the existing traffic information collection methods, development constraints, and the importance of cellular signaling data are summarized, and the architecture of an intelligent transportation system based on cellular signaling data, key technology research progress, and future development direction are reviewed. First, according to its functional planning and development requirements, the architectural design and application framework of the CSD-based urban traffic big data system (C-UTBDS) are proposed. Second, from the perspective of cellular network travel chain construction, the structure of the cellular mobile communication network, travel chain characteristics, and extraction framework are summarized; the noise data of the travel chain, data optimization methods for track vibration, and the road network matching technology, when the travel chain trajectory is integrated with the actual road network, are expounded. Then, considering the needs of urban spatial structure optimization and multi-mode traffic development driven by cellular-network big data, the research status of urban traffic travel characteristics mining is introduced in detail, including population flow monitoring, travel pattern recognition, behavior analysis and prediction. Finally, the technical direction and development trend of future research are highlighted from the aspects of 5G optimization positioning, multi-source data processing and mining, fine-grained travel pattern recognition, and component-based system model architecture construction.