摘要
仿人形机器人各关节的姿态及重心位置是其步态控制研究的关键。为了获得这些姿态数据,设计了一种基于MEMS陀螺仪和加速度计的低成本动态倾角传感器,利用卡尔曼滤波算法将陀螺仪和加速度计输出的数据进行滤波融合,解决了加速度计因机器人的线性加速度而无法准确测量其动态倾角、陀螺仪因存在漂移等而长时间测量精度不高的问题。实际测试结果表明,所设计的传感器能够测量动态倾角并具有较高的精度,可以满足仿人形机器人姿态控制的需要。
The attitude of each joint and centre-of-gravity position are the key factors of humanoid robots gait research.In order to get attitude data,a low cost dynamic tilt sensor based on MEMS gyro and accelerometer was designed.Problems of accelerometer can＇t measure dynamic tilt accurately due to the robot＇s liner acceleration and gyro can＇t do the long time measuring due to the drift error solved by using kalman filter to do the data fusion.Testing results suggested the sensor can measure the dynamic tilt accurately and can be used to do the humanoid robot＇s
出处
《机械设计与制造》
北大核心
2012年第9期141-143,共3页
Machinery Design & Manufacture
基金
西南科技大学青年基金科研项目(11zx3148)
