融合EMA和卡尔曼滤波的MEMS去噪研究与应用  

Research and Application of MEMS Denoising Based on EMAand Kalman Filter

作　　者：岳兴春 彭勇 宋威 黄嘉诚 周钰琛 YUE Xing-chun;PENG Yong;SONG Wei;HUANG Jia-cheng;ZHOU Yu-chen(School of Internet of Things,Jiangnan University,Wuxi 214122,China;School of Artificial Intelligence and Computer,Jiangnan University,Wuxi 214122,China)

机构地区：[1]江南大学物联网学院,江苏无锡214122 [2]江南大学人工智能与计算机学院,江苏无锡214122

出　　处：《仪表技术与传感器》2023年第4期83-86,92,共5页Instrument Technique and Sensor

基　　金：国家自然科学基金项目(62076110)。

摘　　要：针对传统卡尔曼滤波器应用于高灵敏度传感器去噪时的参数初始化问题,通过理论分析了不同测量噪声和系统噪声值对滤波效果的影响,并提出了一种融合EMA和Kalman的滤波算法,以提高对MEMS惯性传感器等高灵敏传感器的滤波效果。仿真结果表明:该算法相对经典卡尔曼算法在收敛前期RMES提升了约13%,在收敛后期提升约10%。并将其应用于MEMS的加速度姿态解算的滤波,结果表明:该算法的滤波效果明显优于经典卡尔曼和互补滤波算法。Aiming at the parameter initialization problem when the traditional Kalman filter is applied to high-sensitivity sensor denoising,the influence of different measurement noise and system noise value on the filtering effect was analyzed theoretically,and a filtering algorithm that combines EMA and Kalman was proposed to improve the filtering effect of highly sensitive sensors such as MEMS inertial sensors.The simulation results show that the algorithm can improve the average RMES by 13%in the initial stage and about 10%in the convergence period compared with the classical Kalman algorithm.And it is applied to the filtering of the MEMS acceleration and attitude calculation.The results show that the filtering effect of the algorithm is obviously better than that of the classical Kalman and complementary filtering algorithms.

关 键 词：高灵敏度传感器 卡尔曼滤波 EMA算法 初始化参数 移动窗口 

分 类 号：TN713[电子电信—电路与系统]