论文部分内容阅读
在基于视觉的机器人轨迹跟踪路径规划中，必须将离散的采样数据复现成原曲线．本文提出一种利用曲线上任意点的切线与水平线的夹角和弧长之间的关系的方法，用圆弧段和直线段来近似拟合二维图线，克服了通常曲线拟合方法不保证高阶导数收敛以及对数据噪声不理想的弱点，实现了曲线的较精确复现，并在自制的直接驱动机器人上进行了实验，取得了满意的效果．
In vision-based robotic trajectory tracking path planning, the discrete sampling data must be reproduced into the original curve. In this paper, a method of using the relationship between the tangent of an arbitrary point on the curve and the horizontal line and the arc length is proposed. The fitting of the two-dimensional graph with circular arcs and straight-line segments approximates the common curve fitting method The convergence of high-order derivatives and the imperfect data noise are guaranteed, and the accurate recurrence of the curves is achieved. Experiments on home-made direct-drive robots are carried out, and satisfactory results are obtained.