摘要
为了提高全自动泊车系统的路径规划质量和跟踪精度,提出了微分平坦路径规划方法和非时间参考滑模的跟踪控制方法。分析了微分平坦系统原理,确定了车辆运动学模型的平坦输出,基于平坦输出量规划了泊车路径。分析了随时间单调递增的非时间参考量,建立了在非时间参考系下的路径跟踪误差模型;在非时间参考系下设计了滑模面和控制律,并证明了Lyapunov意义下的稳定性。经仿真验证可以看出,基于微分平坦原理可以规划出一条最优的泊车路径;同时使用PID控制器和非时间参考滑模控制器对泊车路径进行跟踪,PID控制器的最大纵向误差是滑模控制器的3.57倍,最大方位角误差是滑模控制器的1.67倍,证明了非时间滑模控制器对泊车路径具有更高的跟踪精度。
To improve parking path quality and tracking accuracy of parking system,differential flatness path planning method and non-time reference sliding mode tracking control method are proposed.Differential flatness system principle is analyzed,and flatness output of vehicle motion model is curtained.Parking path is planned based on flatness output.Non-time reference quantity monotone increasing with time is analyzed,and path tracking error model under non-time reference system is built.Sliding mode surface and control law are designed under non-time reference system,and its Lyapunov stability is proved.Through simulation,it can be seen that an optimal path can be planned by differential flatness principle.Parking path is tracked by PID controller and non-time reference controller both.Maximum longitudinal error of PID controller is 3.57 times to non-time reference sliding mode controller,and maximum azimuth angle error of PID controller is 1.67 times to non-time reference sliding mode controller,which proves that non-time reference sliding mode controller possess higher tracking accuracy.
作者
于晓喜
桑建兵
YU Xiao-xi;SANG Jian-bing(Tianjin Transportation Technical College,Tianjin 300110,China;Hebei University of Technology,Tianjin 300401,China)
出处
《机械设计与制造》
北大核心
2020年第9期243-247,共5页
Machinery Design & Manufacture
基金
天津市高等职业技术教育研究会项目(KVIII112)。
关键词
全自动泊车系统
微分平坦
非时间参考
滑模控制
Parking System
Differential Flatness
Non-Time Reference
Sliding Mode Control
作者简介
于晓喜,(1980-),男,天津人,本科,讲师,主要研究方向:汽车运用与维修技术。
