基于改進(jìn)小波變換的MEMS陀螺信號去噪算法

陳光武; 劉孝博; 王迪; 劉射德

doi:10.11999/JEIT180590

基于改進(jìn)小波變換的MEMS陀螺信號去噪算法

doi: 10.11999/JEIT180590

1.
蘭州交通大學(xué)自動控制研究所 ??蘭州 ??730070
2.
甘肅省高原交通信息工程及控制重點實驗室 ??蘭州 ??730070

基金項目: 國家自然科學(xué)基金(61863024, 71761023)、甘肅省基礎(chǔ)研究創(chuàng)新群體計劃(1606RJIA327)、甘肅省自然基金(18JR3RA107, 1610RJYA034)、甘肅省高等學(xué)?？蒲许椖抠Y助(2018C-11)、甘肅省科技計劃資助(18CX3ZA004)

詳細(xì)信息

作者簡介:
陳光武：男，1976年生，教授，研究方向為慣導(dǎo)和組合導(dǎo)航

劉孝博：男，1994年生，碩士，研究方向為慣性導(dǎo)航、傳感器數(shù)據(jù)處理

王迪：男，1991年生，碩士，研究方向為組合導(dǎo)航

劉射德：男，1994年生，碩士，研究方向為視覺導(dǎo)航

通訊作者:
陳光武　cgwyjh1976@126.com

中圖分類號: U666.1; TN911.7
計量
- 文章訪問數(shù): 2921
- HTML全文瀏覽量: 899
- PDF下載量: 108
- 被引次數(shù): 0
出版歷程
- 收稿日期: 2018-06-13
- 修回日期: 2018-12-25
- 網(wǎng)絡(luò)出版日期: 2019-01-04
- 刊出日期: 2019-05-01

Denoising of MEMS Gyroscope Based on Improved Wavelet Transform

1.
Automatic Control Research Institute, Lanzhou Jiaotong University, Lanzhou 730070, China
2.
Gansu Provincial Key Laboratory of Traffic Information Engineering and Control, Lanzhou 730070, China

Funds: The National Natural Science Foundation of China (61863024, 71761023), The Gansu Basic Research Innovation Group Program (1606RJIA327), The Gansu Natural Science Foundation (18JR3RA107 1610RJYA034), Granted by Gansu Provincial Higher Education Research Project (2018C-11), The Gansu Province Science and Technology Plan Funding (18CX3ZA004)

摘要

摘要:
為提高M(jìn)EMS陀螺儀測量精度，抑制測量噪聲對其造成的影響，該文分析了某型號MEMS陀螺儀誤差特性，提出基于遞歸最小二乘法(RLS)多重小波分解重構(gòu)的強追蹤自反饋模型，建立新的軟閾值函數(shù)。由于模型處理后的數(shù)據(jù)帶有部分奇異值，該文提出了一種改進(jìn)的中值濾波算法。對于陀螺儀零偏噪聲問題，提出零偏不穩(wěn)定性抑制算法，并對該算法模型進(jìn)行了詳細(xì)的描述。將某項目研究中列車姿態(tài)測量系統(tǒng)的實驗數(shù)據(jù)應(yīng)用到該算法模型中。測試實驗分為靜態(tài)、動態(tài)兩組，其結(jié)果均表明：該算法減小了信號中的噪聲，有效地抑制了MEMS陀螺儀隨機漂移，提高了姿態(tài)解算的精度?？隙嗽撍惴▽ν勇輧x輸出信號噪聲去除，以及使用精度提升的可行性和有效性。
- MEMS陀螺儀 /
- 小波分解 /
- 姿態(tài)估計
Abstract:
In order to improve the measurement accuracy of Micro Electro Mechanical Systems (MEMS) gyroscopes, the influence of measurement noise on them is suppressed. The error characteristics of a certain type of MEMS gyroscope are analyzed. A strong tracking self-feedback model based on Recursive Least Square (RLS) multiple wavelet decomposition reconstruction is proposed to establish a new soft threshold function. Since the model processed data has partial singular values, an improved median filtering algorithm is proposed. For the problem of gyro zero-bias noise, a zero-bias stability suppression algorithm is proposed. In this paper, the algorithm model is described in detail, and the experimental data of the train attitude measurement system in a project research are applied to the algorithm model. The test experiments are divided into static and dynamic groups. The results show that the algorithm reduces the noise in the signal, suppresses effectively the random drift of the MEMS gyroscope and improves the accuracy of the attitude calculation. The feasibility and effectiveness of this method are affirmed to remove the signal noise of the gyroscope output and improve the accuracy of the use.
- MEMS gyroscope /
- Wavelet decomposition /
- Attitude estimation

HTML全文

圖 1 模型的系統(tǒng)框圖

下載: 全尺寸圖片幻燈片

圖 2 改進(jìn)前后的阿倫曲線圖

下載: 全尺寸圖片幻燈片

圖 3 x 軸原始數(shù)據(jù)濾波處理及相應(yīng)的阿倫方差曲線圖

下載: 全尺寸圖片幻燈片

圖 4 y 軸原始數(shù)據(jù)濾波處理及相應(yīng)的阿倫方差曲線圖

下載: 全尺寸圖片幻燈片

圖 5 z 軸原始數(shù)據(jù)濾波處理及相應(yīng)的阿倫方差曲線圖

下載: 全尺寸圖片幻燈片

圖 6 經(jīng)傳統(tǒng)小波變換和改進(jìn)小波變換處理的姿態(tài)角

下載: 全尺寸圖片幻燈片

圖 7 動態(tài)下各軸角速率的處理結(jié)果

下載: 全尺寸圖片幻燈片

圖 8 動態(tài)下姿態(tài)角結(jié)算結(jié)果

下載: 全尺寸圖片幻燈片

表 1 傳感器性能參數(shù)

	陀螺儀	加速度計	磁力計
測量范圍	±150, ±500, ±1000, ±2000 (°/s)	±2, ±4, ±8, ±16 (g)	±0.6 (mT)
噪聲密度	0.01° (/s·$\sqrt {{\rm{Hz}}} $)	110 (μg/$\sqrt {{\rm{Hzrms}}} $)	48 (nv/$\sqrt {{\rm{Hz}}} $)
敏感度	12.5 mv (/°·s)	1000 (mv/g)	0.1 mv (v·μT)
溫漂	2%	–0.3%/℃	±0.3%
采樣頻率	0.1～200 Hz	0.1～20 Hz	0.1～20 Hz
ARW (°/h^0.5)	1.57	–	–
RRW (°/h^1.5)	600	–	–
BI (°/h)	224.2	–	–

下載: 導(dǎo)出CSV

表 2 兩種小波變換對陀螺儀數(shù)據(jù)處理結(jié)果

算法	坐標(biāo)軸	運行時間(s)	RMS誤差估計	RRW (°/h^1.5)	ARW (°/h^0.5)	BI (°/h)	RR (°/h)
傳統(tǒng)的小波變換	x	26.754976	10.1147	195.2674	0.0301	1.8401	5.3524
	y	28.744975	9.2655	260.4219	0.0283	1.7349	4.5069
	z	27.645963	9.2012	220.3894	0.0117	1.4410	12.7358
改進(jìn)的小波變換	x	26.85396	0.1290	68.6507	0	0	3.0727
	y	28.64576	0.1249	32.9762	0	0	2.3039
	z	27.69872	0.1247	8.6092	0	0	8.7398

下載: 導(dǎo)出CSV

表 3 姿態(tài)解算的MSE誤差估計

坐標(biāo)軸	MSE誤差
坐標(biāo)軸	算法改進(jìn)前	算法改進(jìn)后
z	4.3257×10^–4	1.1512×10^–7
x	8.7754×10^–4	8.5849×10^–7
y	1.5196×10^–4	8.4663×10^–5

下載: 導(dǎo)出CSV

表 4 兩種算法角速率誤差比較數(shù)據(jù)

算法	坐標(biāo)軸	MSE (°/s)	運行時間(s)	MAE (°/s)	ARE (%)
傳統(tǒng)的小波變換	x	0.0421	7.614595	0.0554	11.10
	y	0.0623	8.130619	0.0796	13.41
	z	0.0976	8.647342	0.0842	15.76
改進(jìn)的小波變換	x	0.0999	8.467372	0.0236	8.86
	y	0.0043	7.047250	0.0354	10.87
	z	0.0025	8.021335	0.0416	12.52

下載: 導(dǎo)出CSV

表 5 兩種算法的姿態(tài)角誤差參數(shù)

算法	姿態(tài)角	MSE (°/s)	MAE (°/s)
文獻(xiàn)[20]算法	俯仰角	0.4912	0.4524
	航向角	0.0028	0.1873
	橫滾角	0.0020	0.1171
本文算法	俯仰角	0.2928	0.2360
	航向角	0.0021	0.1354
	橫滾角	0.0014	0.0816

下載: 導(dǎo)出CSV

參考文獻(xiàn)(20)

ZHANG Yanshun, PENG Chuang, MOU Dong, et al. An adaptive filtering approach based on the dynamic variance model for reducing MEMS gyroscope random error[J]. Sensors, 2018, 18(1): 3943–3957. doi: 10.3390/s18113943

XING Haifeng, CHEN Zhiyong, YANG Haotian, et al. Self-alignment MEMS IMU method based on the rotation modulation technique on a swing base[J]. Sensors, 2018, 18(4): 1178–1200. doi: 10.3390/s18041178

WANG Wei and CHEN Xiyuan. Application of improved 5th-cubature kalman filter in initial strapdown inertial navigation system alignment for large. misalignment angles[J]. Sensors, 2018, 18(2): 659–676. doi: 10.3390/s18020659

LI Tao, YUAN Gannan, LI wang, et al. Particle filter with novel nonlinear error model for miniature gyroscope based measurement while drilling navigation[J]. Sensors, 2016, 16(3): 371–385. doi: 10.3390/s16030371

GUO Zhanshe, FU Peng, LIU feng, et al. Design and FEM simulation for a novel resonant silicon MEMS gyroscope with temperature compensation function[J]. Microsyste Technologies, 2018, 24(3): 1453–1459. doi: 10.1007/s00542-017-3524-4

JON O, AIFONSO B, IBAN L, et al. Evaluation of experimental GNSS and 10-DOF MEMS IMU measurements for train positioning[J]. IEEE Transactions on Instrumentation and Measurement, 2018, 6(5): 1–11. doi: 10.1109/TIM.2018.2838799

XIAO Dingbang, XIA Dewei, LI Qingsong, et al. A temperature self-calibrating torsional accelerometer with fully differential configurationand integrated reference capacitor[J]. IEEE Sensors, 2015, 6(7): 1–4. doi: 10.1109/ICSENS.2015.7370428

IGOR P, BROCK B, CAREY M, et al. Towards self-navigating cars using MEMS IMU: Challengesand opportunities[C]. International Symposium on Inertial Sensors and Systems, Moltrasio, Italy, 2018: 1–4.

金靖, 王崢, 張忠鋼, 等. 基于多元線性回歸模型的光纖陀螺溫度誤差建模[J]. 宇航學(xué)報, 2008, 29(6): 1921–1916. doi: 10.387/s100-1328

JIN Jing, WANG Zheng, ZHANG Zhonggang, et al. Temperature errors modeling for fiber optic gyroscope using multiple linear regression models[J]. Journal of Aerospace, 2008, 29(6): 1921–1916. doi: 10.387/s100-1328

DING Jicheng, ZHANG Qian, HUANG Weiquan, et al. Laser gyroscope temperature compensat-i on using modified RBFNN[J]. Sensors, 2014, 14(10): 18711–18727. doi: 10.3390/s141018711

YUAN Guangmin, YUAN Weizheng, LIANG Xue, et al. Dynamic performance comparison of two kalman filters for rate signal direct modeling and differencing modeling for combining a MEMS gyroscope array to improve accuracy[J]. Sensors, 2015, 15(11): 27590–27610. doi: 10.3390/s151127590

ZHA Feng, XU Jiangning, LI JingshuHe, et al. IUKF neural network modeling for FOG temperature drift[J]. Beijing Institute of Aerospace Information, 2013, 24(5): 838–844. doi: 10.1109/JSEE.2013.00097

ZHI S, JACQUES G, MICHAEL J, et al. Low cost two dimension navigation using an augmented Kalman filter/Fast Orthogonal Search module for the integration of reduced inertial sensor system and global positioning[J]. Elsevier, 2011, 19(6): 1111–1132. doi: 10.1016/j.trc.2011.01.001

REN Honglian and PETER K. Investiga-tion of attitude tracking using an integrated inertial and magnetic navigation system for hand-held surgical instruments[J]. IEEE/ASME Transactions on Mechatronics, 2012, 17(2): 210–217. doi: 10.1109/TMECH.2010.2095504

CHEN Xiyuan, XU Yuan, LI Qinghua, et al. Application of adaptive extended kalman smoothing on INS/WSN integration system for mobile robot indoors[J]. Mathematical Problems in Engineering, 2013, 10(10): 1–8. doi: 10.1155/2013/130508

CHU Hairong, SUN Tingting, ZHANG Baiqiang, et al. Rapid transfer alignment of MEMS SINS based on adaptive incremental kalman filter[J]. Sensors, 2017, 17(1): 152–166. doi: 10.3390/s17010152

FENG Yibo, LI Xisheng, and ZHANG Xiaojuan. An adaptive compensation algorithm for temperature drift of micro-electro-mechanical systems gyroscopes using a strong tracking kalman filter[J]. Sensors, 2015, 15(5): 11222–11238. doi: 10.3390/s150511222

BIRSEL A and BILLUR B. Leg motion classification with artificial neural networks using wavelet-based features of gyroscope signals[J]. Sensors, 2011, 11(2): 1721–1743. doi: 10.3390/s110201721

李杰, 曲蕓, 劉俊, 等. 模平方小波閾值在MEMS陀螺儀在信號降噪總的應(yīng)用[J]. 中國慣性術(shù)學(xué)報, 2008, 16(4): 236–239. doi: 10.13695/j.cnki.12-1222/o3.2008.02.03

LI Jie, QU Yun, LIU Jun, et al. Application of modular square wavelet threshold for denoising MEMS-based gyros signal[J]. Journal of Chinese Inertial Technology, 2008, 16(4): 236–239. doi: 10.13695/j.cnki.12-1222/o3.2008.02.03

劉菲, 任章, 李青東. 基于小波方差的MEMS IMU隨機誤差模型間接估計方法[J]. 中國慣性技術(shù)學(xué)報, 2016, 24(1): 77–82. doi: 10.13695/j.cnki.12-1222/o3.2016.01.014

LIU Fei, REN Zhang, and LI Qingdong. Indirect estimation method for random error models of MEMS IMU based on wavelet variance[J]. Journal of Chinese Inertial Technology, 2016, 24(1): 77–82. doi: 10.13695/j.cnki.12-1222/o3.2016.01.014

相關(guān)文章

施引文獻(xiàn)

資源附件(0)

訪問統(tǒng)計