Robotic control of the seven-degree-of-freedom NASA laboratory telerobotic manipulator

A computationally efficient real-time robotic control scheme for the seven-degree of-freedom NASA Laboratory Telerobotic Manipulator (LTM) which utilizes the redundancy of the LTM to avoid joint limits, internal singularities and obstacles is developed. Real-time implementation is essential if the end-effector trajectory is continuously modified based on sensory feedback. Real-time resolution of the redundancy is accomplished in the framework of resolved motion rate control through an efficient implementation of the gradient projection methodwhich does not require the generalized inverse of the Jacobian.

Trajectory planning capability in both Cartesian and joint space is developed. Starting, stopping, and transitions between two path segments are done in a smooth fashion by ensuring continuity in position, velocity, and acceleration. The complete end-effector trajectory does not have to be known, and only the current position and the next desired position are required, which allows immediate modification of the trajectory based on sensory feedback.

A user friendly interface was developed to control the LTM. The interface permits performance criteria and trajectory related information to be added or modified easily. A graphical representation capability showing three orthogonal views of the LTM was developed which permits investigation of the feasibility of control schemes before implementation. A complete set of instructions for the interface is provided.

The robotic control scheme has been shown to be both feasible and effective through real-time graphical simulations demonstrating obstacle avoidance, joint limit avoidance, and singularity avoidance. Integration of the control scheme into the overall system software for the LTM is currently in progress.