摘要
针对避障主动性较差、场景运算量过大,无人驾驶安全与稳定性不理想的问题,提出基于卡尔曼滤波状态预测模型的无人驾驶汽车主动避撞方法。通过构建汽车实际行驶轨迹与理想避撞轨迹之间的偏差目标函数,设立相关约束条件,将问题转化为二次规划问题。使用卡尔曼滤波算法对车辆状态进行预测,结合扩展卡尔曼滤波算法开始滤波递推,得到每一时刻质心侧偏角数据和横摆角速度,完成无人驾驶汽车的主动避撞。实验结果表明:应用所提方法后,质心的纵向位移设置在30 m时,汽车综合行驶平稳性最优;且运用所提方法后的汽车最大侧向位移误差和最大横摆角速度更小,可提升无人驾驶安全与稳定性。
Aiming at the problems of poor active collision avoidance,excessive scene computation,and unsatisfactory safety and stability of driverless driving,an active collision avoidance method of driverless vehicles based on Kalman filter state prediction model is proposed.The problem is transformed into a quadratic programming problem by constructing the deviation objective function between the actual driving trajectory and the ideal collision avoidance trajectory,and setting up relevant constraints.The Kalman filter algorithm is used to predict the vehicle state,and the extended Kalman filter algorithm is combined to start filtering recursion to obtain the sideslip angle data and yaw rate of the center of mass at each time,so as to complete the active collision avoidance of driverless vehicles.The experimental results show that after applying the proposed method,when the longitudinal displacement of the center of mass is set at 30 m,the comprehensive driving smoothness of the vehicle is optimal.And the maximum lateral displacement error and maximum yaw rate of the vehicle after the application of the proposed method are smaller,which can improve the safety and stability of driverless driving.
作者
雍剑书
施文杰
YONG Jianshu;SHI Wenjie(Jiachuang Comprehensive Service Branch of Jiaxing Hengchuang Electric Power Group Co.,Ltd.,Jiaxing 314000,China)
出处
《微型电脑应用》
2025年第8期295-298,共4页
Microcomputer Applications
关键词
无人驾驶汽车
卡尔曼滤波
质心侧偏角
预测模型
主动避撞
路径跟踪
driverless vehicles
Kalman filter
centroid sideslip angle
prediction model
active collision avoidance
path tracking
