A signal processing method to improve the detection accuracy of co gas concentration in TDLAS detection

This paper presents a signal processing scheme that can suppressing background noise arising from laser-induced interference within the detection signal in a tunable diode laser absorption spectroscopy (TDLAS). The constraints imposed on high-precision gas detection in the system by interference fringes were discussed, with particular emphasis on investigating the mechanisms underlying interference fringe generation. In this scheme, the detected signal was processed with a number of “steps” in order to remove part of the signal variation caused by the interference in the TDLAS system. A test set up was used to verify the scheme. The experimental results demonstrate that this algorithm enhances the stability of gas detection. The dispersion of carbon monoxide concentration values decreased by 24.3%. Moreover, this algorithm exclusively addresses the inherent interference noise floor from the laser, and it can be integrated with algorithms aimed at suppressing other interference fringes, thereby enhancing its denoising capability.