基于轨迹预测和模糊分析的商用车盲区防碰撞预警

Trajectory Prediction and Fuzzy Pattern Recognition-based Warning for Commercial Vehicle with Blind Spots

商用车由于其"大、长、高"的特点,车身周围存在诸多视觉盲区,驾驶员难以准确感知环境态势,易导致商用车与周边交通参与者的碰撞.针对事故发生率较高的右转弯工况,提出了一种基于轨迹预测和模糊模式识别的商用车盲区防碰撞预警方法.首先,基于运动学模型和道路边界信息的融合较为精准地预测商用车行驶轨迹;然后融合商用车右转弯过程中A柱盲区和后视镜盲区的空间分布和内轮差效应,构造反映商用车与行人时空间关系的特征指标,用于建立基于模糊模式识别的危险程度分类;最后搭建了商用车右转弯工况盲区防碰撞预警仿真平台并进行相关仿真分析,仿真结果表明,建立的方法可有效实现危险程度的预判.

Commercial vehicles are characterized by its large size,length and height,which bring about many blind spots for drivers.As a result,drivers can’t judge the situation accurately to avoid collisions.A method for collision forecast is proposed for commercial vehicles based on trajectory prediction and fuzzy pattern recognition.First,based on the fusion of kinematics model and road boundary information,the trajectory of commercial vehicles predicts accurately.Then,the characteristics index of the relationship of the spatial distribution between commercial vehicle and the pedestrian are established on the base of the blind spot distribution analysis of the A-column,rear-view mirror and inner wheel difference effect during the right turn of the commercial vehicle.The indexes are helpful for classification of danger level based on fuzzy pattern recognition.Finally,the pre-crash warning simulation platform for the blind spots of commercial vehicle under right-turn condition is built and relevant simulation analysis is carried out.The simulation results show that the proposed method can predict the danger level effectively.