Research on Off-Axis Integrated Cavity Output Spectrum Signal Denoising Based on CSGWO-SVMD-SVD method

Abstract

To mitigate the susceptibility of absorption lines to noise interference during gas measurement using the off-axis integrated cavity output spectroscopy (OA-ICOS) technique, thereby enhancing measurement accuracy, this paper introduces a novel denoising method. This method synergistically integrates Grey Wolf Optimization (GWO) and Cuckoo Search (CS) with Singular Value Decomposition (SVD) and Sequential Variational Mode Decomposition (SVMD). Initially, the optimal solution for the quadratic penalty coefficient in SVMD is ascertained via iterative optimization using the CSGWO algorithm. Subsequently, a circulant matrix is established to extract the singular values of each modal component. A threshold is set to discriminate between noise and useful signals, and the singular values corresponding to noise are nullified. The useful signal components are then reconstructed, yielding the final processed signal. The proposed algorithm was applied to both simulated and experimental target signals and compared with common filtering algorithms such as WT, VMD-WTD, and S-G. In the experimental signal processing results, the signal-to-noise ratio (SNR) of the absorption spectrum signal improved from 21.4 to 39.95, and the correlation coefficient increased from 0.99715 to 0.99946. Results indicate that the proposed algorithm exhibits superior identification and noise suppression capabilities compared to other algorithms. After signal processing using the CSGWO-SVMD-SVD algorithm, the accuracy and stability of signal data detected based on off-axis integration cavity output spectroscopy technology have been greatly improved.