速度测量不确定性影响下的无人车路径跟踪预瞄控制

Preview path following control for autonomous vehicles with inaccurate velocity measurement

预瞄距离是无人车路径跟踪预瞄控制系统设计中的一个关键参数,其选取往往依赖于车辆行驶速度.然而,车速的测量不可避免存在偏差,这必将影响系统的预瞄控制性能.本文针对这一问题,在充分考虑车速测量不确定性对预瞄距离选取影响的基础上,提出一种基于线性参数变化(LPV)系统的鲁棒H∞路径跟踪预瞄控制算法.首先,选取车速和预瞄距离为调度变量,将原系统转化为一种具有可变权重因子的LPV系统.然后,针对无人车系统状态的不完全可测性,设计一种基于观测器的鲁棒H∞控制器.最后,给出保证闭环系统鲁棒H∞性能的凸优化求解条件.仿真结果验证了本文提出控制算法的有效性.

The preview distance is a key parameter for the path following control system of autonomous vehicles,which depends on the vehicle velocity to a great extent.However,there are deviations between the vehicle velocity and its measurements,and the measurement errors on vehicle velocity inevitably deteriorate the path following performance.In this paper,considering the effects of the inaccurate measurement of vehicle velocity on the preview distance,an LPVbased robust H1 control algorithm is proposed to achieve the path following of autonomous vehicles.First,the original nonlinear path following system is formulated as an LPV model with variable weight factors,where the vehicle velocity and the preview distance are chosen as scheduling variables.Then,an observer-based robust H1 path following controller is designed for the case that the system state cannot be measured directly.Finally,the convex conditions for the robust H1 stability of the closed-loop system is provided.Simulation results verify the effectiveness of the control algorithm proposed in this paper.