Research on Obstacle Avoidance of Rope-driven Hyper-redundant Robot Based on Gradient Projection Method

Aiming at the problem that the main body cannot achieve obstacle avoidance when the end of the rope-driven hyper-redundant manipulator performs the main task, this paper proposes an obstacle avoidance algorithm based on the gradient projection method. In this method, the four control points are evenly distributed on the body of the robot arm, and the shortest distance between the control point and the obstacle calculated in real time is used as the optimization index, combined with obstacle modeling and dynamic selection of the amplification factor, to achieve better obstacle avoidance effect. A six-joint twelve-degree-of-freedom rope-driven hyper-redundant manipulator was used for simulation experiments in MATLAB. The results show that the algorithm can realize obstacle avoidance for the rope-driven hyper-redundant manipulator while maintaining a high-end end-effector tracking accuracy, and at the same time the change of the joint angle and the change of the rope length are stable, and there will be no large-scale oscillation.