欠驱动双足机器人的指数稳定周期行走

Exponentially stable periodic walking of under-actuated biped robot

步行机器人环境适应性强、结构复杂、运动控制难.本文研究了基于控制Lyapunov函数和混杂零动态控制的五关节欠驱动双足机器人RABBIT的指数稳定周期步态的行走方法.根据拉格朗日原理建立了机器人腿的混杂动力学模型,并对系统进行了输入/输出线性化处理.结合零动态的思想,构造了控制Lyapunov函数,设计了使混杂模型实现指数稳定的状态反馈控制器,利用限制Poincare的原理,证明了其周期轨道的指数稳定性.仿真结果验证该方法的有效性.

The biped robots have characteristics of strong environmental adaptability,complex structure and difficult motion control.This paper addresses the problem of exponential stable walking of a five-link under-actuated biped robot(RABBIT) through control Lyapunov function and hybrid zero dynamics.The walking system is input-output linearized after the full-order dynamic model is described with Lagrange dynamic equations.Based on hybrid zero dynamics,a state feedback controller is then utilized to realize exponential stability of periodic orbit by designing control Lyapunov function,which is proved to be exponentially stable by restricted Poincare Return Map.The effectiveness of the mentioned method is illustrated by simulations.