动态障碍物意图预测模型构建方法和车辆制动控制方法

The application relates to the field of data processing, and discloses a dynamic obstacle intention prediction model construction method and a vehicle braking control method. The construction method comprises the steps of obtaining Doppler frequency shift signals and point cloud data of dynamic obstacles in front of a visual field of a plurality of vehicles in a historical driving process, extracting micro Doppler features from the Doppler frequency shift signals, extracting motion track time sequence features of the dynamic obstacles from the point cloud data, and jointly training a long-term and short-term memory network according to the micro Doppler features and the motion track time sequence features to obtain a dynamic obstacle intention prediction model. According to the method, the intention prediction model of the dynamic obstacle is obtained through training by using Doppler signals and point cloud data in the driving process of the vehicle, so that the behavior intention of the dynamic obstacle can be accurately predicted, and the safety in the driving process is improved.