The invention belongs to the technical field of pedestrian trajectory prediction, and particularly relates to a no-signal right-turn intersection pedestrian trajectory prediction method based on the game theory, and the method comprises the following steps: S1, obtaining the historical data of pedestrians and vehicles at a no-signal right-turn intersection; s2, analyzing human-vehicle game factors of the non-signal right-turn intersection, and constructing a corresponding human-vehicle game model; s3, inserting the pedestrian-vehicle game model into a preset S-GAN model to obtain an SDG-GAN model for predicting the trajectory of the pedestrian; s4, using the historical data obtained in S1 to train an SDG-GAN model; and S5, using the trained SDG-GAN model to carry out real-time prediction on the pedestrian trajectory of the no-signal right-turn intersection. The method can guarantee the prediction accuracy of the pedestrian trajectory at the no-signal right-turn intersection, guarantees the effectiveness of the auxiliary driving decision at the no-signal right-turn intersection, and gives consideration to the efficiency and safety of the vehicle passing through the no-signal right-turn intersection.
本发明属于行人轨迹预测技术领域,尤其涉及一种基于博弈论的无信号右转交叉口行人轨迹预测方法,包括以下步骤:S1、获取行人及车辆在无信号右转交叉口的历史数据;S2、分析无信号右转交叉口的人车博弈因素,构建对应的人车博弈模型;S3、将人车博弈模型插入到预设的S‑GAN模型中,得到SDG‑GAN模型,用于对行人的轨迹进行预测;S4、使用S1获取的历史数据,对SDG‑GAN模型进行训练;S5、使用训练好的SDG‑GAN模型对无信号右转交叉口的行人轨迹进行实时预测。本发明可以保证对无信号右转交叉口的行人轨迹的预测准确性,保证无信号右转交叉口的辅助驾驶决策的有效性,从而兼顾车辆通过无信号右转交叉口的效率及安全性。
No-signal right-turn intersection pedestrian trajectory prediction method based on game theory
一种基于博弈论的无信号右转交叉口行人轨迹预测方法
2023-06-23
Patent
Electronic Resource
Chinese
European Patent Office | 2024
|European Patent Office | 2024
|Intersection right-turn trajectory planning method in combination with digital traffic rules
European Patent Office | 2024
|