CEEMD阈值滤波在MEMS陀螺信号消噪中的应用

发布时间：2018-02-01 02:46

本文关键词： MEMS陀螺随机噪声互补集合经验模态分解复合指标行人导航　出处：《大地测量与地球动力学》2017年05期 　论文类型：期刊论文

【摘要】：现阶段的经典模态分解MEMS陀螺消噪是一种易造成信号失真缺陷的强制消噪算法。针对该问题,在由本征模态函数样本熵和相似度参数线性组合确定复合指标的基础上,提出基于复合指标筛选本征模态函数的互补集合经验模态分解的阈值滤波降噪算法,根据信号的随机噪声特征参数确定阈值,并对筛选后的本征模态函数进行处理。仿真和实例分析结果表明,该算法的消噪效果较强制消噪滤波有较大提高,使MEMS陀螺信号的零偏不稳定性下降了76.4%,速率斜坡下降了62.3%,均方根误差下降了67.5%,同时能够提高MEMS-IMU在行人导航中的定位精度。
[Abstract]:At present, the classical mode decomposition MEMS gyroscope denoising is a kind of forced de-noising algorithm which can easily cause the signal distortion defect. On the basis of linear combination of sample entropy and similarity parameters of intrinsic modal function, a threshold filtering denoising algorithm based on complementary set empirical mode decomposition (EMD) based on composite index screening intrinsic mode function is proposed. According to the characteristic parameters of random noise, the threshold is determined, and the intrinsic mode function is processed. The simulation and example analysis show that the denoising effect of this algorithm is better than that of the forced filter. The zero bias instability of MEMS gyroscope signal is reduced 76. 4%, the velocity slope is reduced by 62. 3%, and the root mean square error (RMS) error is decreased by 67.5%. At the same time, it can improve the positioning accuracy of MEMS-IMU in pedestrian navigation.
【作者单位】：信息工程大学导航与空天目标工程学院;枣庄市薛城区交通运输局;65631部队;
【基金】：国家自然科学基金(41374027,41674042)~~
【分类号】：P227
【正文快照】： 随着MEMS制造技术和工艺的发展,MEMS陀螺的性能得到了显著的提高,但因MEMS陀螺中所固有的白噪声、闪烁噪声等,使其精度无法胜任长时间导航[1],因此,选择合适的算法去除随机误差显得尤为重要。近年来,小波变换因其良好的时频分析特性和多分辨率能力在工程实践中得到了广泛应用,

【参考文献】