针对四足机器人在非结构化环境下的自适应稳定行走问题,提出一种面向未知复杂地形的四足机器人运动规划方法。采用爬行步态,基于零力矩点(Zero moment point,ZMP)稳定性判据进行在线轨迹规划。通过摆动腿的落地规划和感知策略估计未知...针对四足机器人在非结构化环境下的自适应稳定行走问题,提出一种面向未知复杂地形的四足机器人运动规划方法。采用爬行步态,基于零力矩点(Zero moment point,ZMP)稳定性判据进行在线轨迹规划。通过摆动腿的落地规划和感知策略估计未知地形的参数,实时调整各支撑腿的长度以控制机器人躯体的位置和姿态与当前地形相匹配,实现四足机器人对于未知地形高度和坡度变化的自适应。试验结果表明,四足机器人能够在满足稳定性的前提下,对未知复杂地形具有良好的适应能力,验证了该方法的有效性与可靠性。展开更多
Quadruped robot is considered to be the most practical locomotion machine to negotiate uneven terrain, and shows superb stability during static walking. To improve the ability to go over rough terrain, this paper is f...Quadruped robot is considered to be the most practical locomotion machine to negotiate uneven terrain, and shows superb stability during static walking. To improve the ability to go over rough terrain, this paper is focused on the stable walking and balance control of quadruped robots. 24 kinds of walking gaits are analyzed in order to derive the most stable and smoothest walking gait. Considering the inefficiency to model a terrain by its specified appearance, a uniform terrain model is established and by means of kinematic analysis, a method to adjust the body posture and center of gravity （COG） height is presented. Simulations demonstrate the effectiveness of the proposed meth- od and the improvement of the adaptation of quadruped robots on rough terrain.展开更多
文摘针对四足机器人在非结构化环境下的自适应稳定行走问题,提出一种面向未知复杂地形的四足机器人运动规划方法。采用爬行步态,基于零力矩点(Zero moment point,ZMP)稳定性判据进行在线轨迹规划。通过摆动腿的落地规划和感知策略估计未知地形的参数,实时调整各支撑腿的长度以控制机器人躯体的位置和姿态与当前地形相匹配,实现四足机器人对于未知地形高度和坡度变化的自适应。试验结果表明,四足机器人能够在满足稳定性的前提下,对未知复杂地形具有良好的适应能力,验证了该方法的有效性与可靠性。
基金Supported by the National High Technology Research and Development Program of China(863Program)(2011AA041002)
文摘Quadruped robot is considered to be the most practical locomotion machine to negotiate uneven terrain, and shows superb stability during static walking. To improve the ability to go over rough terrain, this paper is focused on the stable walking and balance control of quadruped robots. 24 kinds of walking gaits are analyzed in order to derive the most stable and smoothest walking gait. Considering the inefficiency to model a terrain by its specified appearance, a uniform terrain model is established and by means of kinematic analysis, a method to adjust the body posture and center of gravity （COG） height is presented. Simulations demonstrate the effectiveness of the proposed meth- od and the improvement of the adaptation of quadruped robots on rough terrain.
