基于横向稳定的减速变道轨迹与跟踪控制  

Decelerating Lane Changing Path and Track Control Based on Lateral Stability

作　　者：徐佳晨 祖炳锋[1,2] 周建伟 董振鹏 XU Jia-chen;ZU Bing-feng;ZHOU Jian-wei;DONG Zhen-peng(Internal Combustion Engine Research Institute,Tianjin University,Tianjin 300072,China;School of Mechanical Engineering,Tianjin University,Tianjin 300072,China;Tianjin Trumpjet Power Technology Co.,Ltd.,Tianjin 300072,China)

机构地区：[1]天津大学内燃机研究所,天津300072 [2]天津大学机械工程学院,天津300072 [3]天津特瑞捷动力科技有限公司,天津300072

出　　处：《公路交通科技》2022年第2期183-190,共8页Journal of Highway and Transportation Research and Development

摘　　要：随着自动驾驶的发展,无人自动驾驶车辆的控制已经成为热点。而无人自动驾驶车辆的安全性和稳定性是重中之重。针对智能驾驶汽车在一定换道距离下保持稳定的减速换道鲁棒性差的问题,提出了一种基于MPC模型预测控制器和比例积分PI控制器相结合双控制器的变道稳定性轨迹确定与跟踪控制方法。为了确定汽车行驶路线,横向与纵向位移分别由5次多项式和匀减速运动公式表示,理论推导了在确定换道距离下满足横向加速度限制的纵向减速度。纵向上通过动力学公式及速度比例积分控制调节制动压强,横向以模型预测控制器输出前轮转角。通过仿真,对不同车速与减速度下的换道时间系数进行标定,得出稳定换道时间与车速和减速度的关系,便于在不同工况中快速切换对应最合适的换道时间系数,防止车辆发生不稳定的动力学表现。最后进行了工况验证。结果表明:在保证横向稳定性的前提下,横向和纵向换道的完成精度分别为99.67%和99.95%,验证了算法的有效性。因此,在自动驾驶汽车轨迹规划和跟踪控制领域中,所提出的双控制器可以提高车辆的控制鲁棒性。With the development of autonomous driving,the control of unmanned autonomous vehicles has become a hot spot.The safety and stability of unmanned autonomous vehicles are the top priority.In view of the poor robustness of smart car maintains stable deceleration under a certain lane-changing distance,a method for determining and tracking the stable lane changing trajectory is proposed based on the dual-controller combining MPC controller and PI controller is proposed.In order to determine the driving route of the vehicle,the lateral and longitudinal displacements are respectively expressed by a 5 th-order polynomial and a uniform deceleration motion formula.The longitudinal deceleration that satisfies the lateral acceleration limit under the determined lane changing distance is theoretically derived.The brake pressure is adjusted by dynamic formula and speed proportional integral(PI) control in the longitudinal direction,and the front wheel turning angle is output by the MPC in the transverse direction.The lane-changing time coefficients at different vehicle speeds and decelerations are calibrated through simulation,and the relationship of stable lane-changing time with vehicle speed and deceleration is obtained,which is convenient to quickly switch the corresponding most suitable lane changing time coefficient under different working conditions to prevent unstable dynamics of the vehicle.Finally,the working conditions are verified.The result shows that under the premise of ensuring the lateral stability,the completion accuracies of the lateral and longitudinal lane changings are 99.67% and 99.95% respectively,which verified the effectiveness of the algorithm.Therefore,in the field of autonomous vehicle trajectory planning and tracking control,the proposed dual-controller can improve the control robustness of the vehicle.

关 键 词：汽车工程 减速变道 横向稳定性 多项式轨迹 模型预测控制器(MPC) 避撞 

分 类 号：U471[机械工程—车辆工程]