【作者】 邹凯； 蔡英凤； 陈龙； 孙晓强；

【Author】 Zou Kai;Cai Yingfeng;Chen Long;Sun Xiaoqiang;Jiangsu University;

【机构】 江苏大学；

【摘要】 针对现有无人车轨迹跟踪研究中将轮胎侧偏角假设在线性区域的不足,提出一种基于增量线性时变模型预测控制的轨迹跟踪方法。在每个控制周期内进行轮胎魔术公式的线性化处理,建立时变轮胎模型,并结合车辆二自由度模型,获得了车辆时变模型,设计增量线性时变模型预测控制器(ILTVMPC),完成了轨迹跟踪,在二次规划求解过程中加入包括控制量和控制增量等约束。利用MATLAB/Simulink平台将该方法与非线性模型预测控制进行仿真对比,结果表明:基于时变轮胎模型的ILTVMPC,不仅在跟踪精度和稳定性上有优异表现,而且计算实时性得到较大幅度提升。更多还原

【Abstract】 To address the deficiency that tire side angles are assumed in the linear region in the self-driving vehicle tracking research,this paper proposes a path tracking method based on incremental linear time-varying model predictive control.The authors linearize the tire magic formula in each control cycle,and establish the tire time-varying model,and combine with the simplified vehicle two-degree-of-freedom model to obtain the vehicle time-varying model.Via designing the Incremental Linear Time-Varying Model Predictive Controller(ILTVMPC),the authors complete the path tracking,and consider the constraints including control amount and control increment in the quadratic programming process.MATLAB/Simulink platform is used to compare ILTVMPC with the Nonlinear Model Predictive Control(NMPC).The results show that ILTVMPC based on the tire time-varying model not only performs excellently in tracking accuracy and stability,but also improves real-time features significantly in computation.更多还原