RRT-based CPC: A configuration planning method for continuum robots using Rapidly-exploring Random Tree algorithm

Abstract

Obstacle-aware configuration control represents a critical challenge in the deployment of continuum robots for advanced applications such as robotic-assisted laparoscopic surgery and intelligent industrial grasping systems. At present, in order to realize the obstacle avoidance function of flexible robots, inverse kinematic calculations are usually unavoidable. The problems of large amount of computation, long solution time, and non-convergence of results make the configuration control for flexible robots still challenging. Most of the current studies use the inverse kinematics calculation of end tracking, and for flexible robots with multiple degrees of freedom, the success rate of obstacle avoidance is low and the computational cost is large. In this paper, a three-segment continuum configuration planning method based on Rapidly-exploring Random Tree (RRT) algorithm is proposed, in which the rough obstacle avoidance path is obtained by RRT algorithm, then the three-segment fitting is carried out by using the second-order Bézier curve, and the length error is evaluated to meet the planning requirements. Experiments such as obstacle avoidance tests, the arrival of target endpoints at different positions and different obstacle environments show that the proposed method can effectively map the feasible solution to the actual configuration. Compared with the inverse kinematics method, the proposed approach improves the success rate of obtaining feasible solutions by at least 14.8% and reduces the solution time by at least 55%. In addition, no prior curvature information and traditional inverse kinematics calculation are needed for the configuration control.