人工肌肉群驱动的柔性仿生机械手臂构型与驱动机理研究

In view of the problem of the existing flexible bionic mechanical arm design, such as simple structure and limited power ability, a novel design method of the flexible bionic mechanical arm joint driven by the artificial muscle group is proposed. Multiple artificial muscles arranged according to some rules are combined into the artificial muscle group, and human upper arm muscle fiber movements are simulated. Configuration design of the artificial muscle group and exploratory research on its driving mechanism for position and force control is presented. The artificial muscle group configuration design and control theory is established. Distribution law of artificial muscle is studied and the multidimensional force control model is established based on kinematics and dynamics analysis. Through space configuration of these artificial muscle groups, global control of muscle groups is conducted and the drive of the whole joint is realized. The compliance and safety of the mechanical arm is ensured, and meanwhile, the overall response speed and load ability is improved. Thereby, artificial muscle flexible drive is further broadened in the applications of the artificial limb, medical rehabilitation, earthquake relief and military equipments. The method has wide engineering application prospect.

针对现有柔性仿生机械手臂设计中存在的结构单一、驱动能力有限等问题，提出人工肌肉群驱动的柔性仿生机械手臂关节设计方法。通过多个人工肌肉束按一定规律排布组合构成人工肌肉群，模拟人体上臂肌纤维的运动；进行人工肌肉群的构型设计，探索实现位置和力控制的人工肌肉群驱动机理，建立一套人工肌肉群构型设计和控制理论。研究人工肌肉束排布规律，根据运动学和动力学分析建立多维受力控制模型，通过这些人工肌肉群在空间上的三维构型，对人工肌肉群进行全局控制，实现整体关节的驱动，保证机械手臂柔顺性和安全性的同时，提高其整体的响应速度和负载能力，从而进一步拓宽人工肌肉柔性驱动器在人工假肢、康复医疗、抗震救灾以及军事装备领域的应用。该方法具有广阔的工程应用前景。

作为一种与生物肌肉特性相似的新型气动元件，气动人工肌肉被广泛应用于仿生机器人的设计与控制研究中。为模拟人体上臂的驱动机理，提高机构的柔顺性，本项目通过多根气动人工肌肉按一定规律排布构成人工肌肉群，模拟人体上臂肌肉的运动，研究气动人工肌肉柔性仿生机械手臂的控制性能。（1）搭建单根气动人工肌肉测试平台，获取气动肌肉静态试验数据，并基于刚度特性系数拟合和BP神经网络两种方法，对气动人工肌肉的力学特性进行分析；（2）搭建气动人工肌肉三自由度关节试验台，基于Lagrange方程建立三自由度关节的动力学模型，并对其进行控制试验研究，为提高仿生机械手臂关节的控制精度提供必要的理论和试验基础；（3）分析人体上臂中骨骼、肌肉及其运动过程中肌肉对骨骼的作用机理，设计人工肌肉群构造的柔性仿生机械手臂，实现机械手臂柔顺性和安全性，并运用拓扑结构分析方法判定机构的合理性；根据运动学和动力学分析建立柔性仿生机械手臂多维受力模型，通过多目标粒子群算法优化柔性仿生机械臂结构参数，并基于遗传算法进行柔性仿生机械臂关节空间轨迹平滑规划；（4）基于以上研究，设计并搭建一套气动人工肌肉群驱动的柔性仿生机械手臂系统；在仿真和试验中，基于计算力矩法设计多种鲁棒性强的控制算法，实现整体关节的有效控制；对机械手臂的控制性能进行分析，对前期的理论研究进行试验验证。本项目的研究为拓宽人工肌肉柔性驱动器的应用提供了一定的理论基础与技术参考。