Multigas TDLAS Detection Based on Variational Mode Decomposition: Frequency Division, Filtering and Interference Noise Removal

Multicomponent gas detection is crucial for industrial and environmental monitoring. In systems based on tunable diode laser absorption spectroscopy (TDLAS), conventional filters with fixed parameters and wide transition bands are unable to effectively separate frequency-adjacent signals or suppress interference fringes, which significantly limits system performance. To address these challenges, we propose the twin-variational mode decomposition demodulation (T-VMD-D) algorithm. This novel approach combines a dual-layer variational mode decomposition architecture with advanced software demodulation techniques, overcoming the inherent limitations of hardware filters. Simulations and experiments demonstrate that the T-VMD-D algorithm achieves significantly higher SNR and robustness in low-interval modulation frequency ranges. At a 100 Hz modulation interval, the algorithm achieves concentration fitting accuracies of 0.9984 (CO₂) and 0.9982 (CH₄), with detection limits improved to 6.74 ppm and 504.81 ppb, respectively. These results validate the excellent sensitivity and immunity of the method proposed in this paper for practical multicomponent gas detection applications. By adaptively tuning mode decomposition parameters, the framework can be generalized to detect additional gas species, enhancing its versatility in multicomponent sensing scenarios.