Self-adaptive CH₄ concentration detection system based on infrared spectrum absorption principle

Abstract

Considering that the noises resulting from low modulation frequency are serious and cannot be totally eliminated by the classic filters, a novel infrared (IR) gas concentration detection system based on the least square fast transverse filtering (LS-FTF) self-adaptive modern filter structure is proposed. The principle, procedure and simulation on the LS-FTF algorithm are described. The system schematic diagram and key techniques are discussed. The procedures for the ARM7 processor, including LS-FTF and main program, are demonstrated. Comparisons between the experimental results of the detection system using the LS-FTF algorithm and those of the system without using this algorithm are performed. By using the LS-FTF algorithm, the maximum detection error is decreased from 14.3% to 5.4%, and also the detection stability increases as the variation range of the relative error becomes much smaller. The proposed LS-FTF self-adaptive denoising method can be of practical value for mid-IR gas detection, especially for weak signal detection.