A Modified Empirical Mode Decomposition Algorithm in TDLAS for Gas Detection

Based on the research of the traditional empirical mode decomposition (EMD) method, we proposed a modified EMD algorithm for the detected signal processing in tunable diode laser absorption spectroscopy. The modified EMD algorithm introduced Savitzky-Golay filtering and cross-correlation operation into the traditional EMD algorithm and reconstructed the signal by using the cross-correlation coefficients effectively. Based on the modified EMD algorithm in theory, the second harmonic component analysis was simulated by comparing with some other filtering algorithms. The experiments system was performed for carbon monoxide (CO) concentration detection. Comparing the sensing performances without and with using EMD-FCR and other filtered methods, the experimental results show that the signal-to-noise ratio of the system was significantly improved from 7.32 to 14.31 dB by EMD-FCR corresponding to one absorption line of CO at 1567.32 nm, leading to the minimum detection limit of 2 ppm. The accuracy and stability of the system are both improved by proposing the modified EMD algorithm.