摘要
在惯性导航系统中,对于动作捕捉时多个惯性传感器的数据实时传输及传输功耗问题,我们在文中提出了蓝牙数据转发装置的设计方案,并建立了惯性系统误差模型,提出了初始对准的方案;基于卡尔曼滤波解决了多传感器的信息融合问题;建立了人体三维棍状模型进行动作实时重现;提出了人体关节活动度与肢体姿态角的转换关系,并以此为姿态数据限幅。
In this article, we focus on the problem of real-time data transmission of multipte inertial sensors in inertial navigation system （INS）. To achieve this goal, we design a bluetooth data- forwarding device and construct an error model of inertial system. Then, an initial alignment method is proposed. Moreover, we settle the problem of multi-sensor information fusion based on Kalman filtering. The motion reconstruction is carried out based on a simplified 3D point-stick model of human body. In addition, we build the conversion relationship between motion range of human joints and body attitude angles, and use this relationship to limit the range of body attitude angles.
出处
《自动化博览》
2017年第3期85-88,共4页
Automation Panorama1
