Litcius/Paper detail

Multi-Joint Bionic Mechanism Based on Non-Circular Gear Drive

Dawei Liu, Tao Zhang, Yuetong Cao

2023Biomimetics10 citationsDOIOpen Access PDF

Abstract

Aiming at the nonlinear expansion/contraction drive problem between different cables in multi-joint cable drive mechanisms, a mechanical drive method based on a non-circular gear drive was proposed, which could replace the servo-sensing control system and minimize the system's complexity and cost. A multi-joint single-degree-of-freedom (DOF) bending mechanism was constructed with several T-shaped components and cross-shaped components. The principle of the multi-joint mechanism driven by non-circular gears was clarified. The corresponding relationships between the joint bending angle, cables' extension/retraction amount and non-circular gear transmission ratio were established. Using the Bowden cable driving, a multi-DOF bending mechanism decoupling scheme was proposed. Considering the adverse effect of non-circular gear hysteresis on the motion of multi-joint mechanisms, a non-circular gear backlash elimination method was proposed. The expression of the backlash of the non-circular gear with respect to the axial movement amount was deduced, which could enable the precise control of the backlash. A two-DOF multi-joint bionic mechanism driven by the non-circular gear was developed. The experimental results show that the mechanism can achieve coordinated bending motion by precisely controlling the line extension/contraction through non-circular gears. This multi-joint bionic mechanism driven by non-circular gears has the characteristics of reliable structure and simple control, and it is expected to be applied to bionic fish and bionic quadruped robots in the future.

Topics & Concepts

BacklashMechanism (biology)Control theory (sociology)Decoupling (probability)EngineeringJoint (building)Structural engineeringComputer scienceMechanical engineeringControl engineeringPhysicsControl (management)Quantum mechanicsArtificial intelligenceSoft Robotics and ApplicationsRobotic Locomotion and ControlRobot Manipulation and Learning