【摘要】 针对视觉伺服机器人手眼定位问题,采用非线性系统理论中的无源化方法,在不需要物体的深度精确值、几何模型及单应性矩阵的计算情况下,设计了摄像机的平移和旋转速度,同时使用自适应控制方法对深度进行估计;基于图像误差实现伺服控制,确保了系统全局渐近可稳定性,该稳定性条件不依赖摄像机的内参数变化。仿真结果表明,利用本方法所设计的控制器,给出了摄像机的速度(包括平移和旋转速度),且能够很好地驱使摄像机由任意初始位置向期望位置运动,实现大范围内渐近稳定。更多还原

【Abstract】 A new approach is proposed to vision-based robot control,which uses the passivity method of the nonlinear system to design the camera translational and rotational velocity without requiring knowledge of geometrical models of the object or precise feature points depth and computing homography matrix.In the case of unknown object depth,an adaptive nonlinear control scheme is used to estimate the points depth.From any initial location the camera can precisely reach its desired position while the system is globally asymptotically stable,and the stability condition does not depend on the variation of the intrinsic parameters of the camera.The simulation shows that our method is effective.The simulation shows that the controller designed with the proposed algorithm can drive the camera precisely to reach its desired position from any initial location and the closed system is globally asymptotically stable.更多还原