Optoelectronics Letters, Volume. 17, Issue 12, 763(2021)

Research on the measurement of CH4 concentration based on dual-band weighted combination model

Honglian LI1,2、*, Wenduo LI1,2, Xiangyu YAN1,2, Heshuai Lü1,2, Fan WANG1,2, and Lide FANG1,2
Author Affiliations
  • 1College of Quality and Technology Supervising, Hebei University, Baoding 071000, China
  • 2National and Local Joint Engineering Center of Measuring Instruments and Metrology Systems, Baoding 071000, China
  • show less

    A detection system based on supercontinuum laser absorption spectroscopy (SCLAS) technology was built up in this study. The absorption spectra of CH4 in different near-infrared bands were measured based on the SCLAS detection system. Under normal temperature and pressure, the absorption spectra of CH4 at different concentrations in the bands of 1 319—1 339 nm and 1 656—1 676 nm were measured. The linear models of CH4 concentrations in different bands were established. The fitting coefficient (R2) of the models were 0.995 7 and 0.996 1, respectively. In order to improve the concentration inversion ability of the model, the weight distribution of each model was determined by R2 and sum of squared error (SSE) to obtain a new combined model. The maximum relative error of the concentration inversion was reduced from 2.3% to 1.3%. The experimental results show the feasibility of applying the SCLAS technology to CH4 concentration measurement, and it also verifies that the dual-band combination model can improve the accuracy and stability of the CH4 concentration measurement.

    Tools

    Get Citation

    Copy Citation Text

    LI Honglian, LI Wenduo, YAN Xiangyu, Lü Heshuai, WANG Fan, FANG Lide. Research on the measurement of CH4 concentration based on dual-band weighted combination model[J]. Optoelectronics Letters, 2021, 17(12): 763

    Download Citation

    EndNote(RIS)BibTexPlain Text
    Save article for my favorites
    Paper Information

    Received: May. 5, 2021

    Accepted: Jul. 1, 2021

    Published Online: Jan. 10, 2022

    The Author Email: Honglian LI (lihonglian@hbu.edu.cn)

    DOI:10.1007/s11801-021-1073-7

    Topics