A strategy for fine motion manipulation of robots using kinematic constraints

This paper examines the problem of controlling a manipulator near a surface. It is assumed that the manipulator must move near or along the surface without contacting the surface. The problem, therefore, incorporates fine motion control and obstacle avoidance. It can occur in applications, such as telerobotic manipulation in space and optical inspection of surfaces. The strategy uses the kinematic constraints that arise from formulating the kinematics and dynamics in the image space of dual quaternion coordinates. Through Lagrange multipliers the constraints are incorporated directly into the dynamics and used to force the manipulator end effector to follow a path along the desired surface. The constraints appear algebraically, and can, therefore, be manipulated directly when developing and implementing strategies for avoiding collisions with the surface. The dynamics are formulated in the operational image space of the manipulator. Thus, the results are easily extended to multiple robot configurations