Utilizing the Entropy Weighting Method to Determine Objective Weights in Robot Trajectory Optimization

Abstract

Optimizing trajectories of collaborative robots is crucial to ensure the efficiency and safety of human-robot interactions in many manufacturing industries. This paper introduced an entropy weighting-based method for determining objective weights in trajectory optimization. Virtual simulations were conducted obtaining datasets of robot trajectory optimization so that, the application scope of the proposed weight determination method could be expanded. The validity of this method was affirmed through comparative simulation experiments. The results demonstrated that in maneuvers in the negative z-axis direction, optimized weights significantly outperformed baseline weights in terms of energy consumption and nominal trajectory tracking, with scores of 0.97 versus 0.68, and 0.82 versus 0.54, respectively. This indicates a better performance than subjective weight setting method in enhancing the performance of collaborative robot trajectory optimization.