葛辰俊,中国农业大学工学院农业工程专业博士研究生。2022年于中国农业大学获得学士学位,曾获得3次校级学习优秀奖学金、1次硕士二等奖学金、1次博士二等奖学金。参与1项国家重点研发计划课题。目前以第一作者在Spectrochimica Acta Part a: Molecular and Biomolecular Spectroscopy发表论文1篇。本人对农业工程领域有浓厚兴趣。
To enhance the universality and accuracy of the near-infrared analysis model for straw aerobic composting across different spectral devices. Corn stover and cow manure were used as raw materials to conduct static trough aerobic composting and membrane-covered aerobic composting. Samples were collected throughout the composting process, and four key parameters moisture content, total nitrogen, total carbon, and organic matter content were analyzed. Near-infrared spectra were collected using benchtop and handheld spectrometers, and corresponding predictive models were developed. Model transfer between the two instruments was achieved using Direct Standardization (DS) and Canonical Correlation Analysis algorithms, enabling better compatibility and data sharing between the benchtop and portable devices. The results demonstrate that both algorithms effectively facilitate the transfer of models for composting process parameters, with the DS algorithm exhibiting superior performance. Specifically, the moisture content and total nitrogen models exhibit good performance after model transfer. The moisture content model of the portable spectrometer achieved an Ratio of Prediction to Devision (RPD) of 2.62, and the total nitrogen model reached an RPD of 2.00. While the total carbon and organic matter content models did not achieve the same high performance levels, there was a significant improvement compared to models without transfer. Their RPD values increased from 0.33 and 0.39 to 1.78 and 1.76, respectively. The two algorithms effectively corrected spectral discrepancies between instruments, mitigating the limitations of instrument variation in near-infrared spectroscopy for composting applications and enhancing the interoperability of different spectrometers. This study provides a novel solution for high-frequency on-site monitoring of the composting process, offering practical strategies for field applications.
葛辰俊,中国农业大学工学院农业工程专业博士研究生。2022年于中国农业大学获得学士学位,曾获得3次校级学习优秀奖学金、1次硕士二等奖学金、1次博士二等奖学金。参与1项国家重点研发计划课题。目前以第一作者在Spectrochimica Acta Part a: Molecular and Biomolecular Spectroscopy发表论文1篇。本人对农业工程领域有浓厚兴趣。
葛辰俊,中国农业大学工学院农业工程专业博士研究生。2022年于中国农业大学获得学士学位,曾获得3次校级学习优秀奖学金、1次硕士二等奖学金、1次博士二等奖学金。参与1项国家重点研发计划课题。目前以第一作者在Spectrochimica Acta Part a: Molecular and Biomolecular Spectroscopy发表论文1篇。本人对农业工程领域有浓厚兴趣。
模型转移是解决模型在不同仪器上通用性问题的总称,其基本原理是尽量减小同一样品在不同仪器上测量信号的不一致性,消除同一仪器间不同时间参数的影响因素[16]。模型转移算法可以分为有标样和无标样2种,其中:有标样算法较为常见[17]。这类算法在使用过程中需要在主仪器和从仪器获取的光谱中选取标样集光谱数据,找出2台光谱仪之间的数学关系。例如直接校正算法(Direct standardization,DS)[18-19]、典型相关分析算法(Canonical correlation analysis,CCA)[20-21]。近年来,在农业领域对模型转移已有很多相关研究。王红鸿等[22]利用DS算法和具有稳定一致光谱信号的波长算法(SWCSS)结合,在纸浆中纤维素含量的检测中,预测均方根误差(RMSEP)降低了1.43%,实现了不同地域、不同批次的光谱仪之间的模型转移。刘智健等[23]在对木质素含量的预测模型中,使用DS算法令2台从仪器的相对偏差误差(Ratio of prediction to deviation,RPD)分别从0.28、0.31提升到了3.39、2.86,实现了不同仪器间的模型转移。Yang等[24]研究了烟草中糖和尼古丁含量的模型转移方法,利用CCA算法在台式和便携式光谱仪之前实现了RMSEP从2.17、0.70降低到2.02、0.39,有效提升了便携式光谱仪的模型性能。
建立模型之前首先需要确定合适的预处理方法,来消除光谱中与样品性质无关的信息,然后采用偏最小二乘回归(Partial least squares regression,PLSR)建立定标模型。图1(b)中3号点的40个样品作为待测样品(外部验证集),这组样品不参与堆肥过程关键参数定标模型的建立,仅用于完成模型转移及验证模型转移效果;其余4个取样点的160个样品用于建立定标模型,为验证模型效果,采用浓度梯度法,按3∶1的比例将160个样品划分为校正集和验证集,其中120个样品作为校正集,40个样品作为验证集。
Al-SuhaibaniN, SelimM, AlderfasiA, El-HendawyS. Comparative performance of integrated nutrient management between composted agricultural wastes, chemical fertilizers, and biofertilizers in improving soil quantitative and qualitative properties and crop yields under arid conditions [J]. Agronomy, 2020, 10(10): 1503
ZengQ D, LiuM F. Application of aerobic composting technology and equipment in agricultural waste recycling[J]. Modern Agricultural Equipment, 2018, 39(2): 53-57 (in Chinese)
[4]
JiZ Y, ZhangL Y, LiuY W, LiX Q, LiZ J. Evaluation of composting parameters, technologies and maturity indexes for aerobic manure composting: A meta-analysis [J]. Science of The Total Environment, 2023, 886: 163929
[5]
YinJ, XieM J, YuX Q, FengH J, WangM Z, ZhangY F, ChenT. A review of the definition, influencing factors, and mechanisms of rapid composting of organic waste [J]. Environmental Pollution, 2024, 342: 123125
[6]
PizzanelliS, CalucciL, ForteC, BorsacchiS. Studies of organic matter in composting, vermicomposting, and anaerobic digestion by 13C solid-state NMR spectroscopy [J]. Applied Sciences, 2023, 13(5): 2900
[7]
ChenL, ChenY N, LiY P, LiuY H, JiangH J, LiH, YuanY, ChenY R, ZouB. Improving the humification by additives during composting: A review[J]. Waste Management, 2023, 158: 93-106
[8]
LuangwilaiT, SidhuH S, NelsonM I. One-dimensional spatial model for self-heating in compost piles: Investigating effects of moisture and air flow [J]. Food and Bioproducts Processing, 2018, 108: 18-26
[9]
LyT B, PhamC D, BuiK D D, NguyenD A K, LeL H, LeP K. Conversion strategies for durian agroindustry waste: Value-added products and emerging opportunities [J]. Journal of Material Cycles and Waste Management, 2024, 26(3): 1245-1263
YuanY, ZhangJ H, LiY Y, GuoE H, ZhangF W. Research progress on aerobic composting of vegetable wastes[J]. Jiangsu Agricultural Sciences, 2022, 50(19): 29-35 (in Chinese)
[12]
GeC J, YangZ L, FanX, HuangY P, ShiZ L, ZhangX T, HanL J. A new spectral simulating method based on near-infrared hyperspectral imaging for evaluation of antibiotic mycelia residues in protein feeds[J]. Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy, 2024, 319: 124536
[13]
LiuM D, LiuJ C, WangQ H, SongP, LiH C, WuS G, GongJ B. Quantitative analysis of low content polymorphic impurities in canagliflozin tablets by PXRD, NIR, ATR-FITR and Raman solid-state analysis techniques combined with stoichiometry [J]. Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy, 2023, 293: 122458
[14]
ZhangS P, TanZ L, LiuJ, XuZ H, DuZ. Determination of the food dye indigotine in cream by near-infrared spectroscopy technology combined with random forest model[J]. Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy, 2020, 227: 117551
[15]
ZhangH, HuX Y, LiuL M, WeiJ F, BianX H. Near infrared spectroscopy combined with chemometrics for quantitative analysis of corn oil in edible blend oil[J]. Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy, 2022, 270: 120841
ZhangJ, CaiW S, ShaoX G. New algorithms for calibration transfer in near infrared spectroscopy[J]. Progress in Chemistry, 2017, 29(8): 902-910 (in Chinese)
XuY Y. Quantitative analysis of methanol in methanol gasoline based on NIR spectroscopy combined with calibration transfer technology [D]. Xi’an: Northwest University, 2022 (in Chinese)