基于主成分分析和SG滤波的光腔衰荡光谱CO气体传感器

郭紫龙; 石程睿; 董媛媛; 张磊; 孙晓园; 孙静静; 周胜

doi:10.37188/CO.EN-2025-0032

基于主成分分析和SG滤波的光腔衰荡光谱CO气体传感器

安徽大学物理与光电工程学院, 光电信息获取与防护技术全国重点实验室, 安徽合肥230601

详细信息

中图分类号: O433.5+4
计量
- 文章访问数: 10
- HTML全文浏览量: 1
- PDF下载量: 1
- 被引次数: 0
出版历程
- 收稿日期: 2025-05-14
- 录用日期: 2025-06-23
- 网络出版日期: 2025-07-22

Cavity ring-down spectroscopy CO gas sensor integrating principal component analysis with savitzky-golay filtering

doi: 10.37188/CO.EN-2025-0032

School of Physics and Optoe-lectronic Engineering, State Key Laboratory of Optoelectronic Information Acquisition and Protection Technology, Anhui University, Hefei 230601, China

Funds: Supported by the National Natural Science Foundation of China (No. 62475001, No. 61905001); the Natural Science Research Project in Universities of Anhui Province (No. 2023AH050088).

More Information

Author Bio:
GUO Zi-long (2000—), male, from Chuzhou, Anhui, is a master’s student who obtained his bachelor’s degree from Anhui University in 2022. He mainly engages in research on cavity ring-down spectroscopy technology. E-mail: 19855101912@163.com

ZHOU Sheng (1900—), male, born in Quzhou, Zhejiang Province, Doctor, Associate Professor, Master’s Supervisor. He obtained a doctorate degree from the Institute of Optics and Electronics, Chinese Academy of Sciences in 2017. He mainly engaged in the research of high-sensitivity laser spectroscopy technology and fiber optic biosensing technology. E-mail: optzsh@ahu.edu.cn

Corresponding author: optzsh@ahu.edu.cn

摘要

摘要:
SG滤波器采用多项式最小二乘近似来平滑数据并估计导数，被广泛用于处理含噪声数据。然而，SG滤波器在数据边界和高频段的噪声抑制能力有限，导致信噪比（SNR）明显降低。为解决该问题，本文提出了一种将主成分分析法（PCA）与 SG滤波协同集成的新方法。这种方法避免了SG滤波较大窗口尺寸带来的过度平滑问题。所提出的PCA-SG滤波算法被应用于基于光腔衰荡光谱（CRDS）的CO气体传感系统。通过与移动平均滤波（MAF）、小波变换（WT）、卡尔曼滤波（KF）和SG滤波器进行对比，验证了PCA-SG滤波算法的性能。结果表明，与所评估的其他算法相比，该算法表现出更优异的降噪能力。衰荡信号的信噪比从11.8612 dB提升至29.0913 dB，提取的衰荡时间常数的标准差从0.037 µs降低至0.018 µs。这些结果表明，所提出的PCA-SG滤波算法有效提高了衰荡曲线数据的平滑度，证明了其可行性。
- 光腔衰荡光谱技术 /
- CO气体传感器 /
- 主成分分析法 /
- SG滤波
Abstract:
The Savitzky-Golay (SG) filter, which employs polynomial least-squares approximations to smooth data and estimate derivatives, is widely used for processing noisy data. However, noise suppression by the SG filter is recognized to be limited at data boundaries and high frequencies, which can significantly reduce the signal-to-noise ratio (SNR). To solve this problem, a novel method synergistically integrating Principal Component Analysis (PCA) with SG filtering is proposed in this paper. This approach avoids the issue of excessive smoothing associated with larger window sizes. The proposed PCA-SG filtering algorithm was applied to a CO gas sensing system based on Cavity Ring-Down Spectroscopy (CRDS). The performance of the PCA-SG filtering algorithm is demonstrated through comparison with Moving Average Filtering (MAF), Wavelet Transformation (WT), Kalman Filtering (KF), and the SG filter. The results demonstrate that the proposed algorithm exhibits superior noise reduction capabilities compared to the other algorithms evaluated. The SNR of the ring-down signal was improved from 11.8612 dB to 29.0913 dB, and the standard deviation of the extracted ring-down time constant was reduced from 0.037 µs to 0.018 µs. These results confirm that the proposed PCA-SG filtering algorithm effectively improves the smoothness of the ring-down curve data, demonstrating its feasibility.
- cavity ring-down spectroscopy /
- CO gas sensor /
- PCA /
- SG filter

HTML全文

Figure 1. Flow chat of PCA-SG algorithm for signal noise reduction.

下载: 全尺寸图片幻灯片

Figure 2. Setup of the CO gas detection system.

下载: 全尺寸图片幻灯片

Figure 3. Simulated absorption coefficient and absorption line intensity of CO based on the HITRAN database.

下载: 全尺寸图片幻灯片

Figure 4. Comparison of the effects of several filtering algorithms on the simulated ring-down signal.

下载: 全尺寸图片幻灯片

Figure 5. Comparison of the residual noise after filtering by several filtering algorithms.

下载: 全尺寸图片幻灯片

Figure 6. Frequency domain distribution of raw noise and filtered residual noise.

下载: 全尺寸图片幻灯片

Figure 7. Comparison of the effects of several filtering algorithms on the measured ring-down signal.

下载: 全尺寸图片幻灯片

Figure 8. Comparison of the residual noise after filtering by several filtering algorithms.

下载: 全尺寸图片幻灯片

Figure 9. (a)CO with different concentrations. (b)The linear fitting of concentrations.

下载: 全尺寸图片幻灯片

Figure 10. (a)Real-time Measured and PCA-SG-Filterd ring-down time. (b)Real-time measured and PCA-SG-Filtered CO concentrations. (c)The Allan deviation plot of the filtered data.

下载: 全尺寸图片幻灯片

Table 1. Comparison of the SNR、RMSE and R² for the filtered signal obtained by different filtering algorithms.

Algorithm	SNR/dB	RMSE	R²
Noise signal	16.2837	5.162×10⁻⁴	0.96205
MAF(3 times)	18.8029	3.332×10⁻⁴	0.97808
WT	21.0154	1.991×10⁻⁴	0.98691
KF	24.3066	9.518×10⁻⁵	0.99374
SG	32.3624	1.451×10⁻⁵	0.99904
PCA-SG	39.9724	2.510×10⁻⁶	0.99983

下载: 导出CSV

Table 2. Comparison of the SNR、RMSE and R² for the filtered signal obtained by different filtering algorithms.

Algorithm	SNR/dB	RMSE	R²
Noise signal	11.8612	6.247×10⁻⁴	0.96797
MAF(3 times)	16.1081	2.482×10⁻⁴	0.97295
WT	18.3120	1.478×10⁻⁴	0.98388
KF	19.7277	1.136×10⁻⁴	0.98760
SG	21.5811	6.903×10⁻⁵	0.99247
PCA-SG	29.0913	1.295×10⁻⁵	0.99858

下载: 导出CSV

参考文献(22)

[1]	BROWN I J, KITIDIS V, REES A P. Simultaneous high-precision, high-frequency measurements of methane and nitrous oxide in surface seawater by cavity ring-down spectroscopy[J]. Frontiers in Marine Science, 2023, 10: 1197727. doi: 10.3389/fmars.2023.1197727
[2]	TANG R W, SONG Y SH, ZHANG H D, et al. A wide dynamic range gas analysis model with deep learning based on cavity ring-down spectroscopy[J]. Sensors and Actuators B: Chemical, 2025, 433: 137575. doi: 10.1016/j.snb.2025.137575
[3]	LI Q Y, LI J, WEI X, et al. An exploratory study on online quantification of isoprene in human breath using cavity ringdown spectroscopy in the ultraviolet[J]. Analytica Chimica Acta, 2020, 1131: 18-24. doi: 10.1016/j.aca.2020.07.056
[4]	AI Y K, LI J, LI Q Y, et al. Cavity ringdown spectroscopy of nitric oxide in the ultraviolet region for human breath test[J]. Journal of Breath Research, 2020, 14(3): 037101. doi: 10.1088/1752-7163/ab8184
[5]	XIONG F, PENG ZH M, WANG ZH, et al. Accurate measurement of trace H₂S concentration based on cavity ring-down absorption spectroscopy under CO₂/CO disturbance[J]. Acta Physica Sinica, 2023, 72(4): 043302. (in Chinese). doi: 10.7498/aps.72.20221851
[6]	LEE J, OH M K. Real-time ultrasensitive detection of ammonia gas using a compact CRDS spectrometer[J]. Applied Optics, 2023, 62(5): 1357-1363. doi: 10.1364/AO.477575
[7]	SONG Y SH, ZHAO J L, ZHANG X N, et al. A CNN-assisted mid-infrared high-sensitivity exhaled ammonia sensor based on cavity ring-down spectroscopy[J]. Sensors and Actuators B: Chemical, 2024, 401: 135071. doi: 10.1016/j.snb.2023.135071
[8]	SHEN M H, HWANG C H, WANG W C. Using higher steps phase-shifting algorithms and linear least-squares fitting in white-light scanning interferometry[J]. Optics and Lasers in Engineering, 2015, 66: 165-173. doi: 10.1016/j.optlaseng.2014.09.004
[9]	DING F. Least squares parameter estimation and multi-innovation least squares methods for linear fitting problems from noisy data[J]. Journal of Computational and Applied Mathematics, 2023, 426: 115107. doi: 10.1016/j.cam.2023.115107
[10]	KALLAPUR A G, BOYSON T K, PETERSEN I R, et al. Nonlinear estimation of ring-down time for a Fabry-Perot optical cavity[J]. Optics Express, 2011, 19(7): 6377-6386. doi: 10.1364/OE.19.006377
[11]	ZHENG J Y, DING F. A filtering-based recursive extended least squares algorithm and its convergence for finite impulse response moving average systems[J]. International Journal of Robust and Nonlinear Control, 2024, 34(9): 6063-6082. doi: 10.1002/rnc.7307
[12]	GOLESTAN S, RAMEZANI M, GUERRERO J M, et al. Moving average filter based phase-locked loops: performance analysis and design guidelines[J]. IEEE Transactions on Power Electronics, 2014, 29(6): 2750-2763. doi: 10.1109/TPEL.2013.2273461
[13]	ZHANG Y J, WU T, ZHANG X ZH, et al. Rayleigh lidar signal denoising method combined with WT, EEMD and LOWESS to improve retrieval accuracy[J]. Remote Sensing, 2022, 14(14): 3270. doi: 10.3390/rs14143270
[14]	LU X Y, LI Y, CHEN X, et al. Discrete wavelet transform assisted convolutional neural network equalizer for PAM VLC system[J]. Optics Express, 2024, 32(6): 10429-10443. doi: 10.1364/OE.516195
[15]	GU L, FEI ZH W, XU X B. Enhancement method of weak Lidar signal based on adaptive variational modal decomposition and wavelet threshold denoising[J]. Infrared Physics & Technology, 2022, 120: 103991.
[16]	ZHOU SH, SHEN CH Y, ZHANG L, et al. Dual-optimized adaptive Kalman filtering algorithm based on BP neural network and variance compensation for laser absorption spectroscopy[J]. Optics Express, 2019, 27(22): 31874-31888. doi: 10.1364/OE.27.031874
[17]	CHANG R J, CHEN ZH, YIN F L. Distributed kalman filtering for speech dereverberation and noise reduction in acoustic sensor networks[J]. IEEE Sensors Journal, 2023, 23(24): 31027-31037. doi: 10.1109/JSEN.2023.3328610
[18]	SCHMID M, RATH D, DIEBOLD U. Why and how savitzky-golay filters should be replaced[J]. ACS Measurement Science Au, 2022, 2(2): 185-196. doi: 10.1021/acsmeasuresciau.1c00054
[19]	ZHANG G SH, HAO H, WANG Y CH, et al. Optimized adaptive Savitzky-Golay filtering algorithm based on deep learning network for absorption spectroscopy[J]. Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy, 2021, 263: 120187. doi: 10.1016/j.saa.2021.120187
[20]	LANJEWAR M G, PANCHBHAI K G, PATLE L B. Sugar detection in adulterated honey using hyper-spectral imaging with stacking generalization method[J]. Food Chemistry, 2024, 450: 139322. doi: 10.1016/j.foodchem.2024.139322
[21]	JANA C, BANERJEE S, MAUR S, et al. Mathematical morphology-based sensing of power system disturbances using PCA-aided support vector machine[J]. IEEE Sensors Journal, 2024, 24(11): 18035-18042. doi: 10.1109/JSEN.2024.3389047
[22]	CHAKRABORTY S, PATRA I, PAL A, et al. Development of a semi-supervised machine learning based noise filter for quantum cascade laser-coupled mid-infrared spectrometer[J]. Infrared Physics & Technology, 2024, 141: 105452.