Infrared gas detection technology based on parameter-tuning stochastic resonance

Authors:
Publisher website
Google Scholar
Abstract
When infrared gas detection technology is applied to monitor gases, the detection accuracy decreases due to the influence of systematic noise. The noises are filtered using traditional denoising methods; however, these methods affect the useful signal. An infrared photoelectric signal is analyzed using the parameter-tuning stochastic resonance method. Based on the parameter normalization frequency and the output SNR tuning bistable system parameters, the system reaches the optimal resonance state, and the signal after denoising is obtained using the Runge–Kutta algorithm. The numerical simulation is conducted on the periodic photoelectric signal with input SNR of 10 dB. It is determined that the method is suitable for the photoelectric signal detection with a frequency from 1 to 10 Hz and that the output SNR is increased to 20 dB in the optimal resonance state. The comparison experiment results show that this method can improve the original signal more effectively than other denoising methods, improving the SNR to 25 dB. In the CO2 standard gas concentration range of 0% to 5%, the maximum relative error of inversion is ±7 % , and the correlation between the inversion result and the standard concentration is 0.9982. This suggests that the proposed method is effective for infrared photoelectric signal detection under strong noise conditions.
Funding Information
  • the national key Research and Development Program of China (2016YFC0201000)
  • Anhui Science and Technology Major Project (15czz04124)