Urban travel behavior analyses and route prediction based on floating car data

The lack of sufficient data is the result of the inherent complexity of gathering and subsequently analyzing route choice behavior, which unfortunately hasn't been revealed much by existing literatures. With the assistance of GIS technology and taxi-based floating car data, the authors found that the majority of urban drivers would not travel along the shortest or the fastest paths. This paper studies the factors that influence commuters' route choice and route switching based on objective real-world observations of travel behavior. Possible factors that may affect driver's route choice are then analyzed and regression methods were introduced to attain if there existing a clear quantitative relationship between drivers' route choice and these factors. The result indicates that such connection is difficult to be established. Consequently, eight scenarios were proposed to quantify the influence of various potential factors. Analysis shows that travel distance, travel time and road preference have comparable higher influence on drivers' route choice. To this end, a new route prediction model is proposed, adopting the road usage as the weight and the shortest route's length and the fastest route's time as the constraints. The proposed model was implemented and validated using the FCD data of Shenzhen, China. The results indicate that by combining the external influence with the driver's personal preference, the predicted travel route has a higher matching ratio with the actual one, which consequently indicates the effectiveness of the model.