Identification of urban taxi passenger hotspots based on massive trajectory data

Trajectory data contains rich information about urban residents' taxi trips, and processing and analyzing the trajectory data to explore the spatial and temporal characteristics of residents' travel behaviors as well as to explore the hotspots of urban taxis is an important basis for improving the efficiency of taxi operation and management level. Firstly, the order information of taxis is identified, and the length and distance distribution characteristics of the orders are mined for each time period of the day. Then the density noise applied spatial clustering algorithm based on grid optimization is proposed to mine the hotspots of urban taxi passengers, and compared with the traditional DBSCAN algorithm, K-distance DBSCAN algorithm and grid density clustering algorithm. The experimental results show that the grid-based DBSCAN algorithm has the highest accuracy and lowest time complexity, and the Davies Bouldin Score (DBI) is reduced by about 35% on average compared with the other three algorithms. The method proposed in this paper has reference value in passenger hotspot mining based on massive taxi trajectory data.