US 12,240,470 B1
Method for driving behavior modeling based on spatio-temporal information fusion
Hong Chen, Changchun (CN); Huihua Gao, Changchun (CN); Ting Qu, Changchun (CN); Yunfeng Hu, Changchun (CN); and Xun Gong, Changchun (CN)
Assigned to Jilin University, Changchun (CN)
Filed by Jilin University, Changchun (CN)
Filed on Oct. 2, 2024, as Appl. No. 18/904,250.
Claims priority of application No. 202411059841.X (CN), filed on Aug. 5, 2024.
Int. Cl. B60W 40/09 (2012.01); B60W 60/00 (2020.01); G06N 3/0442 (2023.01)
CPC B60W 40/09 (2013.01) [B60W 60/0027 (2020.02); G06N 3/0442 (2023.01)] 10 Claims
OG exemplary drawing
 
1. A method for driving behavior modeling based on spatio-temporal information fusion, comprising:
obtaining a training sample set, wherein each training sample in the training sample set comprises spatial sample information, temporal sample information, and a real trajectory sequence of a sample main vehicle;
constructing a driving behavior model, wherein the driving behavior model comprises a spatial information encoding network, a temporal information encoding network, a feature fusion network, and a feature decoding network; the feature fusion network is connected to both the spatial information encoding network and the temporal information encoding network, and the feature decoding network is connected to the feature fusion network;
training the driving behavior model based on the training sample set to obtain a trained driving behavior model; and
determining a future trajectory sequence of a target main vehicle at future time points based on the trained driving behavior model according to spatial information and temporal information of the target main vehicle, wherein the target main vehicle is controlled to travel according to the future trajectory sequence at the future time points; the spatial information comprises a plurality of elements that contain lane markings, a historical trajectory of the target main vehicle, and historical trajectories of background vehicles; and the temporal information comprises the historical trajectory of the target main vehicle and the historical trajectories of the background vehicles.