基于视觉的无人驾驶车自主导航控制器设计

Automatic steering controller design for vision-based unmanned vehicle

以DUTIV-Ⅰ型无人驾驶车为研究对象,建立了基于视觉的无人驾驶车自主导航控制数学模型,应用根轨迹法分析了预瞄距离对该控制系统的影响.设计了由线性二次型最优跟踪控制理论得到的状态反馈控制律和用于补偿参考路径曲率扰动的前馈控制律组成的导航控制器.针对DUTIV-Ⅰ型无人驾驶车横向速度的不可测性,根据Kalman滤波理论构造了横向速度观测器.仿真和实验结果表明,该方法可保证无人驾驶车准确地跟踪参考路径,且具有较强的鲁棒性.

Taking unmanned vehicle DUTIV-Ⅰ as research object,an automatic steering control mathematical model of vision-based unmanned vehicle is established,and the influence of preview distance on automatic steering control system is analyzed using root locus method.An automatic steering controller consisting of state feedback control law obtained by linear-quadratic optimal tracking control theory and feedforward control law used to offset the disturbance of the curvature of reference paths is designed.Aiming at the immeasurable lateral velocity of DUTIV-Ⅰ unmanned vehicle,a lateral velocity observer is developed based on Kalman filter theory.Simulation and experimental results show that the proposed strategy can guarantee high path-tracking accuracy and strong robustness.