Evaluation of advanced trajectory generators and a robust inverse kinematic solver with a 7-DOF industrial robot

An optimized trajectory of multi-axis robotic manipulators is important and directly influences the performance of the executing task. In manipulators of seven or more axis, this optimization involves considerations regarding continuity of position and velocity in both Cartesian and joint space coordinates. In this paper, two trajectory generation algorithms are applied on a 7-axis manipulator for performance evaluation. These two trajectory generation algorithms evaluated here, namely the Cubic Spline and the Modified Tension Spline, are different in their mechanism and simplicity, and their evaluation will yield conclusions and guidelines which can be readily applied in the automation industry. Two application tasks are designed to evaluate the combined performance of the trajectory generators and the inverse kinematics (IK) solver; a triangular wave trajectory in a specific plane and a three-level shelf reaching trajectory. The results show that, near the vicinity of joint limits and singularities, a Cartesian pose which undershoots and overshoots will most certainly cause convergence errors in IK solvers. In addition, both a Cartesian space trajectory and pose optimization can potentially decrease the convergence of inverse kinematic solvers. Finally, the inclusion of low pass filters in the IK output will most certainly reduce power costs, increase the lifetime of motors and boost the stability of the accuracy and precision of robot position control.