基于一种智能辅助步行手杖的工程化欠驱动柔性腿式机器人步行控制研究

It is currently a hot topic and difficult problem in the robotics field to achieve efficient and fast walking of engineered legged robots as animals by using the combination of compliant legs and locomotion control techniques. This project proposes to develop a novel intelligent walking-assist stick for man (i.e., a compliant one-legged robot) aiming at the shortcomings of the existing engineered legged robots such as large mass, complex structure and insufficient robustness of control strategy, on the theretical basis of structural design and locomotion control of the underactuated comliant legged robots. The global time-varying dynamics model corresponding to different walking patterns (gaits and attitudes) is to be built up and identified by viewing the user and the intelligent stick as a “generalized compliant legged robot” . Meanwhile, the double-layer locomotion control framework and related techniques of the underactuated legged robot are then to be developed with empahsis to realize the control goals for the energy injection and maintenance of the man-machine system, the periodic planar and three-dimensional walking with variable patterns, and the aperiodic walking including posture balance recovery, etc. This study will promote a solution of the robust and adaptive walking control of the underactuated legged robot in the real world, which not only provides the theoretical basis and practical example for the continuous research on the legged robot locomotion control, but also gives the technical support for the innovative walk-assist tools in an aging society.

利用具有柔性关节的腿部与运动控制技术的结合，使工程化腿式机器人实现类似动物的高效、快速步行，是当前机器人领域的一个热点和难点问题。本项目基于欠驱动柔性腿式机器人运动控制的理论基础，针对现有工程化腿式机器人质量大、结构复杂以及控制技术应用鲁棒性能不足等缺陷，提出研发一种新型辅助人类步行的智能手杖（单腿柔性关节机器人），将使用者和手杖视为“广义柔性腿式机器人”建立并辨识对应不同行走模式（姿、步态）的整体时变动力学模型，并重点发展欠驱动腿式机器人双层运动控制框架与相关技术，实现人机系统能量注入与保持、可变行走模式的平面和三维周期性步行及姿态平衡恢复等非周期性步行的控制目标。本研究将推动解决在现实环境中欠驱动柔性腿式机器人进行鲁棒自适应步行控制的关键技术问题，成果不仅能为腿式机器人运动控制可持续性研究提供理论依据与实践范例，而且也能为老龄化社会所需的新型智能辅助步行工具发展提供技术支持。

用具有柔性关节的腿部与运动控制技术的结合，使工程化腿式机器人实现类似动物的高效、快速步行，是当前机器人领域的一个热点和难点问题。本项目基于欠驱动柔性腿式机器人运动控制的理论基础，针对现有工程化腿式机器人质量大、结构复杂以及控制技术应用鲁棒性能不足等缺陷，提出研发一种新型辅助人类步行的智能手杖（单腿柔性关节机器人），将使用者和手杖视为“广义柔性腿式机器人”建立并辨识对应不同行走模式（姿、步态）的整体时变动力学模型，并重点发展欠驱动腿式机器人双层运动控制框架与相关技术，通过与下肢外骨骼机器人结合实现控制目标。本研究将推动解决在现实环境中欠驱动柔性腿式机器人进行鲁棒自适应步行控制的关键技术问题，成果不仅能为腿式机器人运动控制可持续性研究提供理论依据与实践范例，而且也能为老龄化社会所需的新型智能辅助步行工具发展提供技术支持。