融合卡尔曼改进小波的姿态传感器滤波方法  

Attitude Sensor Filtering Method Based on Kalman’s Improved Wavelet

作　　者：王静静 秦刚[1] 李腾 WANG Jingjing;QIN Gang;LI Teng(School of Electronic Engineering Information Engineering,Xi’an Technological University,Xi’an 710021,China)

机构地区：[1]西安工业大学电子信息工程学院,西安710021

出　　处：《自动化与仪表》2023年第9期110-114,119,共6页Automation & Instrumentation

摘　　要：为解决XGZT-XXX型姿态传感器输出噪声大且受复杂环境干扰的问题,该文提出基于时间序列建模的卡尔曼联合改进小波去噪的滤波方法。该方法通过建立ARMA时间序列模型作为卡尔曼滤波器的输入完成首次滤波;而后使用改进的小波模糊阈值去噪将滤波后的信号分解后的高、低频同时进行阈值处理,高频分量使用小波模糊阈值去噪,低频使用最小二乘平滑滤波,从而实现减小噪声影响的目的。引入Allan方差辨识陀螺仪的噪声结合信噪比均方差作为评价指标,通过实验结果表明,该文方案量化噪声减少63.17%,角度随机游走减小81.57%,零偏不稳定性减少87.59%,角速率随机游走减少81.64%,速率斜坡减少77.80%,信噪比增加7.7 dB,均方差下降65.45%,可有效提高XGZT-XXX型姿态传感器的稳定性和测量精度。In order to solve the problem of large output noise and complex environmental interference of XGZT-XXX attitude sensor,a filtering method for Kalman joint improved wavelet denoising based on time series modeling is proposed.In this method,the first filtering is completed by establishing the ARMA time series model as the input of the Kalman filter.Then,the improved wavelet fuzzy threshold denoising is used to decompose the filtered signal at the same time for high and low frequency threshold processing,the high-frequency component uses wavelet fuzzy threshold denoising,and the low frequency uses least squares smooth filtering.So as to achieve the purpose of reducing the influence of noise.The experimental results show that the quantization noise of the proposed scheme is reduced by 63.17%,the angular random walk is reduced by 81.57%,the biased instability is reduced by87.59%,the angular rate random walk is reduced by 81.64%,the rate slope is reduced by 77.80%,the signal-to-noise ratio is increased by 7.7 dB,and the mean square deviation is reduced by 65.45%.It can effectively improve the stability and measurement accuracy of XGZT-XXX attitude sensor.

关 键 词：MEMS器件 ARMA模型 小波模糊阈值 ALLAN方差 信噪比 

分 类 号：TP212[自动化与计算机技术—检测技术与自动化装置]