The Analysis of Community Travel Behaviors Based on Big Trajectory Data 2020-01-5239
The rapid growth of internet-based trips has greatly changed the travel habits of urban community residents and urban traffic patterns. Currently, many residents prefer to travel by hailing a vehicle online for its convenience, which generates a great deal of data for intelligent traffic systems and can help practitioners and researchers better investigate urban traffic problems. In this paper, the community travel behaviors were investigated using the large trajectory datasets generated by online ride hailing services. First, the popular communities are selected based on Density-Based Spatial Clustering of Applications with Noise (DBSCAN) method. Then, a spatial travel regularity of communities is analyzed, followed by a rank analysis of communities. Besides, the house price is also considered in this paper. Finally, an origin/destination (O/D) prediction method is proposed based on bidirectional long-short term memory (BiLSTM) and attention mechanism. To improve the superiority of prediction method, we have compared our method with some baselines, such as, basic BiLSTM, Support Vector Regression (SVR), and Autoregressive Integrated Moving Average model (ARIMA). The predicted results show that our proposed O/D prediction method is better than other baselines. From the travel behavior analysis and O/D prediction, we can see that the analysis can characterize the community travel behaviors overall.