| 基于自适应模型预测的智能汽车横向轨迹跟踪控制 |
| |
| 引用本文： | 张志达,郑玲,张紫微,李以农,梁艺潇.基于自适应模型预测的智能汽车横向轨迹跟踪控制[J].中国公路学报,2022,35(7):305-316. |
| |
| 作者姓名： | 张志达 郑玲 张紫微 李以农 梁艺潇 |
| |
| 作者单位： | 重庆大学 机械传动国家重点实验室, 重庆 400044 |
| |
| 基金项目： | 国家自然科学基金项目(51875061)；重庆市技术创新与应用发展专项(cstc2019jscx-zdztzxX0032);重庆市研究生科研创新项目(CYB19063) |
| |
| 摘 要： | 当路面附着情况和车辆行驶状态不断变化时，基于恒定侧偏刚度的模型预测控制(MPC)不能考虑轮胎非线性特性的影响，难以保证车辆轨迹跟踪的适应性。为此，提出一种考虑轮胎侧向力计算误差的自适应模型预测控制(AMPC)，以提高智能汽车在不确定工况下的轨迹跟踪性能。分析了路面附着系数和垂向载荷对轮胎侧向力的影响，基于平方根容积卡尔曼滤波(SCKF)算法，设计了利用侧向加速度和横摆角速度作为测量变量的前后轮胎侧向力估计器。利用轮胎侧向力线性计算值与估计值的差值计算得到侧偏刚度修正因子，设计了前后轮胎侧偏刚度的自适应修正准则，进而提出了一种基于时变修正刚度的AMPC控制方法。基于CarSim与MATLAB/Simulink联合仿真和硬件在环测试平台，对AMPC控制的有效性和实时性进行了验证。研究结果表明：在不同的路面附着情况和车辆行驶状态下，AMPC控制都能够降低横向位置偏差和航向角偏差，有效提高车辆的轨迹跟踪精度，其控制效果明显优于基于恒定侧偏刚度的标准MPC控制。尤其在低附着工况下，标准MPC控制会因为线性轮胎力的计算误差过大而导致车辆在轨迹跟踪时严重失稳，而AMPC控制通过估计轮胎力修正侧偏刚度依然能够保证车辆稳定有效的跟踪参考轨迹。所提出的AMPC控制在保证控制精度的同时具有良好的实时性，对智能汽车控制系统的设计与优化具有重要参考价值。
|
| 关 键 词： | 汽车工程 轨迹跟踪控制 自适应模型预测 轮胎侧向力估计 平方根容积卡尔曼滤波 |
| 收稿时间： | 2020-12-25 |
Lateral Trajectory Tracking Control of Intelligent Vehicles Based on Adaptive Model Prediction |
| |
| ZHANG Zhi-da,ZHENG Ling,ZHANG Zi-wei,LI Yi-nong,LIANG Yi-xiao.Lateral Trajectory Tracking Control of Intelligent Vehicles Based on Adaptive Model Prediction[J].China Journal of Highway and Transport,2022,35(7):305-316. |
| |
| Authors: | ZHANG Zhi-da ZHENG Ling ZHANG Zi-wei LI Yi-nong LIANG Yi-xiao |
| |
| Affiliation: | State Key Lab of Mechanical Transmissions, Chongqing University, Chongqing 400044, China |
| |
| Abstract: | Model predictive control (MPC) based on constant cornering stiffness does not consider the influence of the nonlinear characteristics of tires under constantly changing road adhesion conditions and vehicle driving states, thereby making it difficult to guarantee the adaptability of vehicle trajectory tracking. Therefore, this study proposes an adaptive model predictive control (AMPC) method considering the tire lateral force calculation error to improve the trajectory tracking performance of intelligent vehicles under uncertain conditions. The influence of the road adhesion coefficient and vertical load on the tire lateral force was analyzed. Based on the square-root cubature Kalman filter algorithm, a lateral force estimator for the front and rear tires was designed, by considering the lateral acceleration and yaw rate as measurement variables. The correction factors of the cornering stiffness were calculated based on the difference between the calculated and estimated tire lateral forces, the adaptive correction criterion of the front and rear tire cornering stiffness was established, and an AMPC method based on the time-varying modified stiffness was proposed. The performance of the AMPC method was verified and compared with that of the standard MPC method based on the CarSim and MATLAB/Simulink joint simulation and hardware in the loop test platform. The results indicate that AMPC can reduce the lateral position deviation and heading angle deviation and improve the tracking accuracy of vehicles under different road adhesion conditions and vehicle driving states. The control effect of AMPC is significantly better than that of the standard MPC with a constant lateral stiffness. Particularly under the low-adhesion condition, the standard MPC causes serious instability of the vehicle in the trajectory tracking because of the large calculation error of the linear tire force, whereas the AMPC enables the vehicle to track the trajectory stably and effectively by estimating the tire force to correct the cornering stiffness. The proposed AMPC method yields a good real-time performance while ensuring control accuracy and is expected to serve as an important reference for the design and optimization of intelligent vehicle control systems. |
| |
| Keywords: | automotive engineering trajectory tracking control adaptive model prediction tire lateral force estimation square-root cubature Kalman filter |
|
| 点击此处可从《中国公路学报》浏览原始摘要信息 |
|
点击此处可从《中国公路学报》下载全文 |
