Near-Time-Optimal Control of a Robot Manipulator using new Approximation Techniques

We present a new approach for determining a time-optimal trajectory for a robot manipulator with actuator torque bounds. We find an approximation to a globally optimal trajectory which passes between two configurations will avoiding obstacles. Our algorithm accomplishes this by by dynamically generating and searching a finite graph. Our algorithm is guaranted to produce a trajectory that performs within a user-specified factor of a time-optimal safe trajectory.