A robust and efficient approach for the time optimization of path constrained motions of robotic manipulators incorporating actuator torque and jerk constraints

摘要

A method is proposed for minimizing the time of motion along a predefined geometrical movement in configuration space, with all joints as parametrized functions of one common scalar lambda . The principles of time minimization are described and used as the basis of an approach in which the trajectory is built from elementary sets of predefined time-dependent functions. A priori knowledge about the robot's dynamical behavior is used to keep the necessary torque and jerk values of all actuators within predefined limits. Apart from the suitability of the representation for storage and reproduction of position, velocity, and acceleration of all joints, extra robustness and a reduced calculation load are achieved by using this explicit formulation together with a specially tuned two-stage optimization algorithm.