In this paper, a multi-band fusion model to improve the performances of the super-continuum laser absorption spectrometer (SCLAS) of CO 2 was proposed and demonstrated. Various concentrations of CO 2 were measured by the super-continuum laser in the wavelength of 1425–1445 nm, 1565–1585 nm, and 1595–1615 nm, respectively, at 295 K and 1 atm. The method for derivation of CO 2 concentration using the integrated area of spectrum peaks is proposed. Linear models of the CO 2 concentration and the integrated area of the absorption peaks in different bands were established, which achieves R 2 of 0.9947, 0.9937, and 0.9824, respectively. The measurement accuracy with the models is evaluated with the parameter of relative analysis error ( RPD ), which results in a of more than 2, indicating reasonable prediction ability of the models. In order to improve the accuracy of the single model, the models of the three bands are weighted and fused based on R 2 and RMSE , respectively. One of the fusion models reduces the prediction error and improves the accuracy of the single model effectively by decreasing the maximum relative error from 3.4 to 1.2% for a single model. The experimental results show that SCLAS can measure the CO 2 concentration under different environments. The multi-band fusion model proposed here is feasible for CO 2 measurement, which provides a new idea and new method for the detection of gases.

中文翻译：

---

基于多波段融合模型的CO2浓度测量研究

在本文中，提出并论证了一种用于提高 CO 2 超连续谱激光吸收光谱仪 (SCLAS) 性能的多波段融合模型。通过超连续谱激光器在 1425-1445 nm、1565-1585 nm 和 1595-1615 nm 波长下，在 295 K 和 1 个大气压下测量各种浓度的 CO 2。提出了利用谱峰积分面积推导CO 2 浓度的方法。建立了CO 2 浓度和不同波段吸收峰积分面积的线性模型，R 2 分别达到0.9947、0.9937和0.9824。模型的测量精度以相对分析误差（RPD）为参数进行评价，a大于2，说明模型的预测能力合理。为了提高单个模型的精度，三个频段的模型分别基于R 2 和RMSE 进行加权和融合。其中一种融合模型通过将单个模型的最大相对误差从3.4%降低到1.2%，有效地降低了预测误差，提高了单个模型的精度。实验结果表明，SCLAS可以测量不同环境下的CO 2 浓度。本文提出的多波段融合模型对于CO 2 测量是可行的，为气体检测提供了新思路和新方法。其中一种融合模型通过将单个模型的最大相对误差从3.4%降低到1.2%，有效地降低了预测误差，提高了单个模型的精度。实验结果表明，SCLAS可以测量不同环境下的CO 2 浓度。本文提出的多波段融合模型对于CO 2 测量是可行的，为气体检测提供了新思路和新方法。其中一种融合模型通过将单个模型的最大相对误差从3.4%降低到1.2%，有效地降低了预测误差，提高了单个模型的精度。实验结果表明，SCLAS可以测量不同环境下的CO 2 浓度。本文提出的多波段融合模型对于CO 2 测量是可行的，为气体检测提供了新思路和新方法。