An Gradient-based Continuous-Time Trajectory Optimization method for Real-Time Collision Avoidance on collaborative robot

Abstract

Collision avoidance is an essential consideration in the growing application of human-robot interaction (HRI) and robot-robot interaction (RRI). Based on predicting the trajectory of other agents, a continuous-time trajectory optimization method based on the gradient for real-time collision avoidance of manipulators is proposed in this paper. The algorithm constructs a gradient field in joint space. It optimizes trajectory by using convex hull characteristics of the B-spline so that the robot can avoid dynamic obstacles and generate a smooth and dynamic feasible trajectory in joint space. The proposed collaborative robot (Co-robot) trajectory planning framework can be calculated in real-time and verified by simulation.