基于MEMS的室内定位误差修正方法研究

根据室内惯性定位存在误差累积的特点,建立广义似然比检测的方法进行零速度检测,利用Kalman滤波对检测到的"零速度"时刻进行零速修正(zero velocity update,ZUPT),从而有效降低系统累积误差。但行人行走过程中存在的无效振动,导致测得的加速度和角速度数据中出现明显的噪声,这对长时间定位精度产生较大的影响。文中提出在利用Kalman滤波进行误差校正之前首先采用Butterworth低通滤波滤除加速度和角速度数据中由无效振动引起的高频部分,即噪声部分,从而消除行人运动过程中的无效振动对定位精度的影响。

A study of indoor positioning error correction method based on MEMS

According to the accumulative error characteristics of the indoor inertial navigation ,this paper exploits a generalized likelihood ratio detector to execute zero velocity detection .And then according to the moment of ‘zero velocity ’ detected , Kalman filter is adopted to correct the accumulative errors (zero velocity update , ZUPT ) , thereby effectively reducing the system errors . However , because of the influence of invalid vibration during the pedestrians walking ,it leads to the obvious noise in acceleration and angular velocity data ,which has great influence on positioning accuracy for a long time .To address this concern ,the paper adopts the order Butterworth low‐pass filter to filter out noise (high frequency part) caused by invalid vibration among the navigation data before the zero velocity is updated (ZUPT) by Kalman filter .