We use a hy-brid visual representation to minimize the dimensionof state space and to overcome pose ambiguity. 3D pa-rameters are estimated directly from the image plane,thus not requiring a calibration step. The designed con-trol law ensures global asymptotic stability in the senseof Lyapunov, while in the case of bounded uncertain-ties the closed-loop system behavior is characterizedby a global attractor. Experimental results validate thetheoretical framework both in terms of system conver-gence and control robustness. As future work, we aregoing to include the dynamics of manipulator in thevisual model and considering applications in mobilerobotics, where additional constraints are due to thepresence of a nonholonomic mobile base.Appendix AProof of Lemma 1. Consider the following Lyapunovfunction candidate:V.x x x; t/ D 12Q x x xT Q x x x; (A.1)which is positive definite and radially unbounded. Thetime derivative of V.x x x; t/, using Eq. (16), results:P V.x x x; t/ D Q x x xT P Q x x x D Q x x xT.B.x x x/cV V V cno − P x x x.d//D−6 XiD1ki
jQ xi j; (A.2)which is negative definite 8ki > 0;i D 1;::: ,6.
愿景提供了一套强大的感官流程的机器人移动非结构化环境中,
因为它允许的非接触测量外界和增加任务准确性。在操作初期基于视觉的机器人
系统。使用的方法的开环的方式称为静态查找然后移动。精度可大幅度提高
通过位置环闭合基于视觉反馈。论文网
这种方法的主要目的,称为视觉伺服,是控制位置相对于所述机器人到目标对象或一组的目标特征。图1示出的一般的块的视觉伺服系统。该系统接受两个输入对任务的描述要执行的(参照输入),以及对象运动(干扰输入)。视觉分析提供了视觉的说明,作为与当前相机位置的环境;描述由控制器块进行处理,其产生的摄像机运动。
视觉伺服系统可以被分类为适当的视觉伺服系统或动态的外观和移动系统。在第一种情况下,所述关节力矩是基于视觉控制器的输出,而在第二种情况下,控制器的输出是参考拧螺丝输送到内部位置/速度环。
视觉伺服系统英文文献和中文翻译(5):http://www.751com.cn/fanyi/lunwen_69953.html