Spatial and Temporal Distribution Model for Travel Origin-Destination Based on Multi-Source Data

Scientifically grasping the spatial and temporal distribution of road traffic demand is a prerequisite for formulating traffic congestion countermeasures. This paper analyzes the relationship between built environment attributes and travel origin-destination (OD) under the specific spatial structure of the city, and provides guidance for decision-makers to solve traffic congestion problems. Based on multi-source data, the paper uses the Dirichlet-multinomial regression (DMR) model to discover the functional zones and recognize the spatial structure of the city. Based on the obtained spatial structure, the spatial autoregressive (SAR) model is then applied to establish the relationship between the urban built environment attributes and the spatial and temporal distribution of the travel OD. Finally, using the data of the downtown area of Chengdu as an example, the model and method are verified and analyzed.