摘要
当前主流的仿人机器人都采用ZMP(zero moment point)理论作为稳定行走的判据。实时ZMP点落在支撑足与地面接触形成的多边形支撑区域内是仿人机器人实现稳定步行的必要条件。因此实现仿人机器人在复杂现实环境中稳定行走,必须要求机器人足部感知系统提供足够丰富的地面环境信息,从而可以准确获取支撑区域的形状以实现基于实时ZMP点的稳定控制。文中将柔性阵列力传感器应用于仿人机器人足部感知系统,提出了获取仿人机器人支撑区域形状的方法,而且通过实验验证了其可行性。
The concept of zero moment point ZMP,as a dynamic equilibrium criterion,has been so far extensively adopted in the mainstream humanoid robots for real-time and dynamic control of stable walking or running.The necessary condition for biped walking is： real-time ZMP point of robot falls on the polygonal supporting area formed by the supporting foot contacting with ground.Therefore,to make humanoid robot walk stably in complex reality environment,the robot foot perception system should provide rich information about ground environment,so that we can obtain the shape of supporting area accurately to achieve stability control based on real-time ZMP point.The flexible force sensor array is used in robot foot perception system and a method of obtaining the shape of humanoid robot’s supporting area is proposed,and its feasibility is verified by experiment.
出处
《自动化与仪表》
北大核心
2010年第10期1-4,共4页
Automation & Instrumentation
基金
国家高技术研究发展计划(863计划)面上项目(2008AA04Z205)
国家高技术研究发展计划(863计划)重点项目(2008AA04206(01))
关键词
柔性阵列力传感器
仿人机器人
足部感知系统
零力矩点
支撑区域
flexible force sensor array
humanoid robot
foot perception system
zero moment point（ZMP）
supporting area
