摘要
探讨了智能手机惯性传感器校正的九参数模型的6种形式,利用模量方程构建椭球模型,并通过多项式分解得到类似线性化的迭代模型,进而提出了一种似线性模量标定方法。利用该方法进行手机传感器多位置标定实验时,无需严格控制传感器标定位置或姿态,便可获得传感器的校正参数。以三星S3手机采集的加速度计和陀螺仪数据进行标定实验,计算结果与分析表明,所提方法适用于智能手机传感器的快速标定。
Six forms of nine-parameter model are discussed in this paper, and the quasi-linear vector module method for the error calibration of MEMS inertial sensors is presented based on nine-parameter model. The estimation procedure of the new method is very simple, iteration based on polynomial de- composition, which is completely different with first-order Taylor iteration. Multi-position calibration experimental re- suits show that the iteration number of the quasi-linear vec- tor module method is much less than five, the accuracy of the novel method is equal with that of the existed component method, and the provided method has no need to control sensor attitude using strict test platform in the existed compo- nent method.
出处
《测绘地理信息》
2016年第6期20-24,共5页
Journal of Geomatics
基金
大地测量与地球动力学国家重点实验室开放基金资助项目(SKLGED2014-3-2-E)
国家自然科学基金青年项目(41204011)
国家高技术研究发展计划(863计划)资助项目(2013AA12A201)~~
关键词
智能手机
惯性微机电传感器
误差标定
九参数模型
似线性模量法
smart phone
MEMS inertial sensors
error calibration
nine-parameter model
the quasi-linear vector mod- ule method
