Abstract
Near-infrared spectroscopy (NIRS) is an efficient method for detecting the content of carbon and nitrogen in many materials, which solves the problems of the time-consuming and high-cost traditional chemical analysis method. To quickly detect the carbon-nitrogen ratio (C/N) for the anaerobic fermentation (AF) feedstock using NIRS, a genetic simulated annealing algorithm (GSA) is presented based on a genetic algorithm combined with a simulated annealing algorithm. By combining GSA with backward interval partial least squares (BiPLS), we construct a BiPLS-GSA algorithm to optimize the characteristic wavelength variables of NIRS; this algorithm significantly reduced the number of wavelength variables involved in modeling and effectively improved the detection accuracy and efficiency of the model. The determination coefficients, root mean squared error, mean relative error (MRE) and residual predictive deviation for the validation set in the BiPLS-GSA regression model were 0.9067, 7.6676, 5.5274%, and 3.5626, respectively. Meanwhile, compared to the entire spectrum model, the MRE was decreased by 16.54% in the BiPLS-GSA-based model. The research in this paper improves the adaptability of the prediction model based on optimizing sensitive wavelength variables for C/N, which provides a new way for rapid and accurate measurement of the C/N of AF feedstock.
© 2019 Optical Society of America
Full Article | PDF ArticleMore Like This
Lili Xu, Jinming Liu, Chunqi Wang, Zhijiang Li, and Dongjie Zhang
Appl. Opt. 62(11) 2756-2765 (2023)
Changhao Bao, Changhao Zeng, Jinming Liu, and Dongjie Zhang
Appl. Opt. 61(19) 5790-5798 (2022)
Na Wang, Longwei Li, Jinming Liu, Jianfei Shi, Yang Lu, Bo Zhang, Yong Sun, and Wenzhe Li
Appl. Opt. 60(15) 4282-4290 (2021)