Online optimization enhanced closed-loop control of multi-section continuum robots

Abstract

Due to the inherent characteristics of continuum robots (high flexibility, multiple degrees of freedom), controlling the continuum robots safely and precisely in practical applications has always been a challenging task. In this paper,a real-time kinematic closed-loop controller that optimizes the step length to boost control performance is proposed. Initially, a differential-based generalized inverse kinematics solution is formulated to resolve the DOF coupling in twin-pivot continuum robots that intertwined two DOFs in one joint. Subsequently, an adaptive online optimization strategy utilizing the algorithm of Particle Swarm Optimization (PSO) is proposed to refine the controller, overcoming the limitations of traditional Jacobian-based approaches. This novel method innovatively decouples control direction and step length, optimizing safety and efficiency. Comparative simulations and tracking tests confirm the controller's superior precision and efficiency, with an average accuracy of 0.33 %, a 35 % enhancement over the Jacobian controller, thus facilitating the broader application of multi-section continuum robots.