The activities of freight vehicles are significantly influenced by broader transportation market demands and operational strategies. Given that effective operational activities for trucks primarily occur during stopping periods, it is essential to conduct further research into truck parking behavior. To better understand the spatial–temporal patterns of truck operations in China. This paper proposes an improved Long Short-Term Memory (LSTM) multivariate prediction model to forecast stopping cities and stopping duration using trajectory data from Chinese freight management platform. The model integrates historical and current truck stopping and driving features, effectively capturing spatial and temporal patterns in the data. Based on parking behavior indicators, the truck population is categorized into four groups, and the potential reasons for varying prediction accuracy across trucks are explored. Compared to traditional LSTM and Back Propagation Neural Network (BP) models, the proposed method improves the prediction accuracy of stopping cities for truck samples by 1.3% and 2.0%, respectively, and increases the accuracy of stopping duration predictions by 6.3% and 8.1%. The findings indicate that trucks with higher spatial stability and temporal continuity tend to exhibit higher prediction accuracy. These results offer new insights for truck mobility prediction modeling and behavior pattern research and provide theoretical support for truck route planning.
A Freight Truck Stopping Behavior Prediction Approach Based on Trajectory Dataset
Int. J. ITS Res.
International Journal of Intelligent Transportation Systems Research ; 23 , 1 ; 475-488
01.04.2025
14 pages
Aufsatz (Zeitschrift)
Elektronische Ressource
Englisch
Truck trajectory data , Stopping behavior characteristics prediction , Machine learning , Long Short-Term Memory , Attention Mechanism Engineering , Electrical Engineering , Automotive Engineering , Robotics and Automation , Computer Imaging, Vision, Pattern Recognition and Graphics , Civil Engineering , User Interfaces and Human Computer Interaction
A Freight Truck Stopping Behavior Prediction Approach Based on Trajectory Dataset
Springer Verlag | 2025
|British Library Conference Proceedings | 1992
|