Design, Modelling and Control of a Continuum Manipulator Based on Multilayer Planar Springs

Abstract

There is a surge of research interest in the field of “continuum robotics”. Robots created under this paradigm offer many advantages and represent unique features in terms of flexibility, dexterity, safety and weight reduction. In the thesis, a novel continuum manipulator that integrates multiple layers of compliant planar springs is introduced – a structure that provides several notable advantages over existing designs. It possesses precise linear large-displacement motion and demonstrates effectively decoupling bending from contraction and thus reduces the uncontrolled compression when generating normal deflections; besides, an enlarged workspace of the end-effector is achieved by varying the length of the continuum manipulator via contraction. The mechanics of the proposed continuum manipulator is investigated. An analytical method is provided to study the compliance characteristics of planar springs and derive the unified compliance matrix to represent the force-deflection relationships, allowing an accurate motion prediction. Differences of the compliance characteristics with respect to design variations of planar springs are discussed. An analysis regarding behaviours of the full continuum manipulator is given. According to the constantcurvature approximation, two kinematic models corresponding to three-tendon-driven and single-tendondriven continuum manipulators are presented. This modelling methodology permits closed-form kinematics and also facilitates the derivation of differential kinematics and real-time control.
In view of the model’s complexity and uncertainty of the continuum manipulator, a fuzzy control approach is implemented for autonomous execution of end-effector motion tasks. The system state-space model is constructed using the general continuum manipulator kinematics with the constant-curvature assumption. The fuzzy controller is designed utilizing state-feedback control techniques. Thus, this control methodology enables a low-computation solution to this motion control problem without the need for continuously updating the Jacobian of the continuum manipulator. Besides, compared to traditional Jacobian-based controllers that suffer from model inaccuracies, the fuzzy control exhibits superior performances with respect to a specified cost function.

Date of Award	2016
Original language	English
Awarding Institution	King's College London
Supervisor	Thrishantha Nanayakkara (Supervisor), Hongbin Liu (Supervisor) & Kaspar Althoefer (Supervisor)