绳驱动软体操作臂的建模与控制技术研究

Research on Modeling and Control of A Tendon-driven Soft Manipulator

研究了一款绳驱动软体操作臂的建模与控制技术。基于分段恒曲率模型,研究了驱动空间与关节空间之间、关节空间与任务空间之间的映射和逆映射关系。使用MATLAB软件进行了软体操作臂的工作空间分析和运动性能仿真。根据运动学原理,搭建实验样机平台,通过伺服电机控制驱动绳的长度。在样机平台上进行了软体操作臂的运动控制实验,结果显示分段恒曲率模型不足以准确地控制软体操作臂。在单变量条件下和多变量条件下,分别使用曲线拟合和BP神经网络的方法,对软体操作臂的运动进行了补偿控制。

This paper studied the modeling and control of a tendon-driven soft manipulator. The soft manipulator was made of flexible silicone with molds, and driven by 3 tendons. Based on the piecewise constant curvature （PCC） model, the kinematic analysis was carried out, including the mapping relationship among actuator space, configuration space and task space. MATLAB software was used for workspace analysis of the soft manipulator and simulation of kinematic property. According to the principle of kinematics, an experimental prototype platform was set up, and the lengths of tendons were controlled by servo motors. A set of motion control experiments were conducted on the prototype platform, and results showed that the PCC model cannot control the soft manipulator accurately enough. Under the conditions of single variable and multiple variables, curve fitting and BP neural network were respectively used to compensate and control the motion of the soft manipulator.