基于AHRS反馈的仿人机器人步行稳定控制  被引量:3

Stabilizing control of humanoids'walking based on AHRS feedback

在线阅读下载全文

作  者:张继文[1,2,3] 刘莉[1,2,3] 陈恳[1,2,3] 

机构地区:[1]清华大学机械工程系,北京100084 [2]精密超精密制造装备及控制北京市重点实验室,北京100084 [3]摩擦学国家重点实验室,北京100084

出  处:《清华大学学报(自然科学版)》2016年第8期818-823,共6页Journal of Tsinghua University(Science and Technology)

基  金:国家自然科学基金资助项目(51175288;51105219)

摘  要:以实现仿人机器人步行稳定控制为目的,为克服零力矩点(ZMP)反馈控制算法不能补偿较大冲击干扰以及反射响应方法打断当前步行任务的缺点,以低成本微机电系统(MEMS)传感器及方向余弦矩阵描述的补偿算法构成航姿参考系统。提出了由姿态转角和角速度的线性组合表达的步行稳定性判据,并实现了躯干姿态校正的比例—微分控制器。基于仿人机器人全方位步态规划方法,调整下一步步幅和步频为行为方式,实现了步行稳定控制。样机实验结果表明:在原地踏步和稳定前进等情况下,机器人受到冲击干扰以致倾斜角度达20°时,仍能够不打断全方位步行且有效避免倾覆。An attitude and heading reference system (AHRS) feedback system was established using micro-electro-mechanical system (MEMS) sensors with low cost and the compensation algorithm described by the direction cosine matrix to realize the stabilizing control of humanoids' walking, being aimed to overcome the defect of incapability to compensate the large disturbance when using the zero moment point (ZMP) feedback control algorithm and the shortcoming of the current walking task interruption when using the reflex response control method. The walking stability criterion expressed by the linear composition of posture angles and angular velocities was also proposed, with a proportion derivative controller for torso pose adjusting also implemented. Walking stabilization control was realized by the action style of tuning the next stride and walking frequency using the omni directional walking gait planning method. Experimental results with a physical robot show that the robot can effectively avoid omnidirectional walking task interruption and fall over even if disturbed by an impulse to reach the maximum inclination of 20° when marching on the spot and walking forward steadily.

关 键 词:仿人机器人 航姿参考系统 步行稳定 全方向步行 

分 类 号:TP242.6[自动化与计算机技术—检测技术与自动化装置]

 

参考文献:

正在载入数据...

 

二级参考文献:

正在载入数据...

 

耦合文献:

正在载入数据...

 

引证文献:

正在载入数据...

 

二级引证文献:

正在载入数据...

 

同被引文献:

正在载入数据...

 

相关期刊文献:

正在载入数据...

相关的主题
相关的作者对象
相关的机构对象