基于狩猎型尺蠖软体机器人冗-欠驱动与柔性控制研究

The soft robot can change its shape and size arbitrarily, and has wide application prospect in the field of unstructured, dynamic and unknown complex environment. A hybrid actuation and dynamic variable structure control strategy based on the adaptive camouflage, high-speed movement, large-deformation, precise positioning and capture, multi-state snatch of harvester caterpillars are proposed in this project. The main research contents are included in this project as following: (1) mathematical description and establishment of nonlinear dynamic model based on analysis of the hunting behavior of harvester caterpillars, and building soft robot prototype by using multi-material embedded combination mechanism. (2) the integrated hybrid actuated mode of the local redundant-the overall under actuation by using of pneumatic, smart memory actuator and piezoelectric actuator; the law of vibration transmission and dissipation, and building vibration suppression mechanism based on honeycomb, soft materials and 3D printing. (3) the proposed multi-mode flexible control mechanism, including movement trend control based on fuzzy control, variable stiffness / variable damping control on-line adjusted by motion parameter, and multi-mode displacement-force adaptive dynamic switching strategy combined with sliding mode control, adaptive control and neural network based on hunting. The project will provide the theoretical guidance and technical basis for the application of the soft robot.

软体机器人可任意改变自身形状与尺寸，在非结构化、动态、未知复杂环境领域中有广泛的应用前景。本项目针对狩猎型尺蠖捕食过程中自适应伪装、高速运动、弯曲大变形、精确定位及捕捉、多态抓举等多边复杂性，提出了基于狩猎型尺蠖软体机器人的混合驱动及动态变结构控制策略，内容包括：（1）分析狩猎型尺蠖捕食行为的运动机理并进行数理描述，建立非线性动力学模型，使用多材料嵌入组合机制构建软体机器人本体。（2）采用气动、记忆金属、压电等实现软体机器人的局部冗余-整体欠驱动的一体化混合驱动模式；研究捕食过程中振动的传递与耗散规律，利用蜂窝、软材料、3D打印等构建减振抑振机制。（3）提出多模式柔性控制机理，即模糊控制实现运动趋势控制，构建运动参数在线调整的变刚度/变阻尼控制机制，结合滑模控制、自适应控制与神经网络实现基于狩猎型软体机器人的多模位移-力自适应鲁棒动态切换策略，为软体机器人的应用提供理论指导与技术基础。

软体机器人可任意改变自身形状与尺寸，在非结构化、动态、未知复杂环境领域中有广泛的应用前景。本项目针对狩猎型尺蠖捕食过程中自适应伪装、弯曲大变形、定位及捕捉等多边复杂性，通过3D打印模具、软体浇筑及内嵌技术构建了基于气动网络、形状记忆金属的多款软体机器人本体，软体机器人的抓取夹持控制策略，内容包括：（1）分析狩猎型尺蠖捕食行为的运动捕食机理，借助气动网络构建包络式抓取软体机器人、尺蠖运动软体爬行机器人、冗余驱动软体机器人，采用有限元法借助Abaqus软件完成软体机器人力学分析。（2）探索智能材料（如压电）的非线性，建立迟滞模型，通过自适应滑膜控制算法、（非线性）PID算法实现对迟滞非线性的补偿；利用蜂窝、软材料、3D打印等技术构建减振抑振结构。（4）利用气动压力传感器、弯曲传感器，基于Arduino软件设计软体机器人运动控制策略，通过PID控制、非线性PID，切换控制实现软体机器人弯曲及力的控制。