To improve both efficiency and safety of automatic driving in complex traffic scenarios, autonomous vehicles need to have the ability to predict the future trajectories of vehicles. The interaction among vehicles in real traffic scenarios makes the trajectory prediction of vehicles challenging. By utilizing the historical trajectory of the targeted vehicle and the surrounding environment information, this paper proposes a coupling LSTM model in order to effectively predict the future trajectories of vehicles. The proposed model predicts the observable motion intentions of vehicles, and builds grids for the targeted vehicles to extract the implied space and intention information of the neighboring vehicles. Experiment results verify the proposed model in contrast to other basic models in real traffic scenarios.
Intelligent Prediction of Mobile Vehicle Trajectory Based on Space-Time Information
01.04.2019
1003285 byte
Aufsatz (Konferenz)
Elektronische Ressource
Englisch
Vehicle trajectory prediction method based on time-space fusion attention
Europäisches Patentamt | 2025
|Space-time attention LSTM vehicle trajectory prediction method based on position-speed
Europäisches Patentamt | 2023
|Attention Based Vehicle Trajectory Prediction
IEEE | 2021
|