A QUATERNION-DRIVEN DEEP LEARNING-BASED NOVEL APPROACH FOR MOBILE AND LOCOMOTIVE ROBOT PATH PLANNING AND MOTION PREDICTION

Abstract

In this study, I address the locomotive-robot dilemma in movement task sequences. Our method combines geometric motion planning and locomotion prediction using quaternions and deep learning architecture. This is comparable to human motion prediction. I begin by developing a collision-avoidance-based motion planning method. Then, using transformer deep learning, I anticipate robot locomotion. I used simulation to demonstrate my findings.