Determination of Via-point Using Threshold-based Segmentation Algorithm for Joint Space Trajectory Profile

Abstract

Via-point is a mid-way point between the starting and stopping trajectory points. Finding the via-point from a complex trajectory point data series is challenging due to unknown start and stop points. Therefore, this paper uses the segmentation algorithm to segment the joint space trajectory profile to find the via-point and number of phases in the joint trajectory profile. Our algorithm deals with multiple joint robot configurations to ensure the number of phases is the same for all joints. The algorithm finds the standard deviation, $\sigma$ of each joint profile, and selects the highest value referred to as the most dominant joint, Jd, during movement execution. Then segment the Jd based on direction change as a reference to all other joints. We show that the algorithm can locate a via-point to reduce the complexity of robot motion. It shows that the algorithm can produce the same number of segments for the repetitive joint motion.