面向超动态柔性运动的六足机器人自由步态规划与耦合运动机理研究

The hexapod robot has a broad application prospect in various fields. Achieving ultra-dynamic flexible movement has become an urgent need to promote its application. However, limited by the joint flexibility and motion flexibility problem, the existing hexapod robots are difficult to improve movement performance essentially. In this project, these issues are to be investigated. Firstly, based on the research on flexibility generate mechanism of biological joint, a new variable stiffness flexible joint based on potential energy principle and an active/passive flexibility control method are proposed to achieve bionic flexibility. Then, reveal the gait generation mechanism of creatures, and the basic frame and methods of free gait planning are proposed based on discretization. On this basis, the reinforcement learning free gait planning method and adaptive free gait planning method are researched by imitating the learning behavior and adaptive mechanism of biological gaits to satisfy the needs of robot flexible movement. Finally, the coupling motion mechanism of the creatures is researched and the efficient body motion planning method and foot trajectory planning method are explored. A kind of multi-legged coordinated control method fusing joint flexibility control is proposed to realize the robot ultra-dynamic flexible movement in complicated dynamic environment and improve the robot overall movement performance essentially. This project will perfect the theoretical system of ultra-dynamic flexible motion and promote the development and application of the hexapod robot, so it has important academic significance and application value.

六足机器人具有广阔的应用前景，实现超动态柔性运动已成其推广应用的迫切需求，然而受限于“关节柔性与运动柔性”问题，现有六足机器人难以在运动性能上获得质的提升。针对上述问题，本项目首先基于生物关节柔性生成机理研究，构建基于势能调控的新型变刚度仿生柔性关节，提出非预测外力下主/被动相融合的关节柔性控制方法，实现关节仿生柔性；其次，揭示生物步态生成机理，创新基于离散化思想，提出自由步态规划的基本框架与方法，模仿生物步态的学习行为与自适应机制，研究增强学习与自适应自由步态规划方法，满足机器人柔性运动需求；之后，研究生物复杂行为耦合运动机理，探索高效的机体运动与足端轨迹规划方法，融合关节柔性控制，提出六足机器人多足协调控制方法，实现复杂动态环境下机器人超动态柔性运动，本质提升机器人整体运动性能。本项目研究及其成果对完善六足机器人超动态柔性运动理论体系，推动六足机器人发展应用具有重要学术意义和应用价值。

六足机器人因具备丰富的步态和冗余的肢体结构，在国民经济与国防建设关键领域有着广阔的应用前景。然而，受限于“关节柔性与运动柔性”问题，现有六足机器人难以在运动性能上获得质的提升，根源上制约了六足机器人的推广应用。项目面向六足机器人超动态柔性运动需求，聚焦机器人亟待解决的“关节柔性与运动柔性”两大瓶颈问题，系统开展了仿生柔性关节、自由步态规划、多足协调控制等相关研究。研究借鉴生物关节柔性生成机理，构建了基于势能调控的新型变刚度仿生柔性关节，考虑关节负载的时变特征与非预测性，提出了非预测外力下主/被动相融合的关节柔性控制方法，初步解决了机器人关节柔性问题；研究借鉴生物自由步态生成机理，创新基于离散化思想，提出了自由步态规划的基本框架与方法，模仿生物步态的学习行为与自适应机制，建立了增强学习与自适应自由步态规划方法，基本攻克了机器人运动柔性难题；研究借鉴生物复杂行为耦合运动机理，提出了基于运动相对性的机体运动规划方法和基于参数化思想的足端轨迹规划方法，基于机体与足端运动解耦研究，建立了六足机器人多足协调控制方法，重点突破了机器人柔顺控制瓶颈。项目初步揭示了生物关节柔性生成机理、步态生成机理及复杂行为耦合运动机理，重点突破了仿生柔性关节设计与主/被动柔性控制、自由步态规划和多足协调控制等关键技术，构建了一种基于势能调控的新型变刚度仿生柔性关节，研发出多款六足机器人原理样机，并从新的视角提出了一套较为完备的六足机器人步态规划与运动控制理论方法，初步实现了复杂动态环境下六足机器人超动态柔性运动。项目研究及其成果对丰富完善六足机器人超动态柔性运动理论体系，推动六足机器人发展应用具有重要意义。