基于加权时变卡尔曼滤波的惯性测量单元设计  

Design of inertial measurement unit based on weighted time-varying Calman filter

作　　者：伍婷婷 单奇[1] 

机构地区：[1]西南交通大学机械工程学院,四川成都610031

出　　处：《计算机工程与设计》2018年第1期134-139,共6页Computer Engineering and Design

摘　　要：针对加速度计和陀螺仪数据融合姿态测量方法在载体运动加速度较大时得到的角度误差逐渐增大的现象,提出一种基于加权时变的卡尔曼滤波算法。利用衡量运动加速度相对大小的权重参考概念,设置加速度计测量角度在观测量中所占的比重,使系统能够根据实际情况改变对加速度数据的参考程度。基于Arduino平台,选用MPU6050惯性测量传感器进行实时测量数据获取姿态信息的实验研究,实验结果表明,改进后的算法可以有效减小由于系统运动加速度导致的测量误差,获得更准确的姿态角数据。In view of the phenomenon that the angle error of accelerometer and gyroscope data fusion attitude measuring method increase gradually when the carrier's movement acceleration is larger,Calman filtering algorithm based on time-varying weighting was put forward.The weight reference concept used to measure acceleration's relative size was used,the measurement proportion of accelerometer measurement angle of observed quantity was set,which made the system change the reference level of acceleration data according to the actual situation.Based on the Arduino platform,using MPU6050 as the inertial measurement sensor,the attitude information of data acquisition of real-time measurement was studied.Experimental results show that the improved algorithm can effectively reduce the measurement error caused by the acceleration of the system motion,and obtain more accurate attitude angle data.

关 键 词：数据融合 惯性测量 姿态测量 权重参考 卡尔曼滤波 

分 类 号：TP212[自动化与计算机技术—检测技术与自动化装置]