基于滚动时域估计方法的智能小车循迹控制

本资料包含pdf文件1个，下载需要1积分

Aiming at the path following of a robotic car, a control strategy was proposed based on moving horizon estimation（MHE）to improve its accuracy and rapidity. A planar dynamic mathematical model of the robotic car was established from the view of overlooking. To achieve widening range of vision and accurate path tracking, the path target was obtained by using a CCD camera fixed on the top of the car as the basis for the study of MHE control strategy. The optimal controller design was obtained by dealing with a fixed length of time optimization problem with online constrains. Therefore, the calculate complex was reduced. Computing horizontal and vertical ordinate of the path target simultaneously, an arc-shaped target with 25 cm ahead was tracked in the experiment. The main results show that the proposed strategy is reasonable, feasible, and lays a foundation for further improving performance of the path tracking issue.针对智能车路径跟踪的准确性与快速性要求，提出了基于滚动时域估计（MHE）方法的优化控制策略。首先从俯视的角度，建立了智能车的平面动态数学模型；为了拓宽智能车的视野，提高其路径识别的准确性，通过利用车载CCD摄像设备获取目标道路信息，并以此为基础提出了滚动时域循迹控制策略；基于滚动时域估计方法，将对小车控制量的估计问题转化为固定时长的优化问题，避免了计算量随时间不断增大，使得在线处理系统约束问题成为可能；试验中，通过获取的目标道路信息同时计算横纵坐标，并控制小车跟踪前方25cm处开始的一段弧形轨迹。试验及研究结果表明，该设计方案是合理并可行的，为进一步提高智能车循迹控制性能奠定了基础。