基于实验型和工业型近红外光谱仪的木片材性预测建模对比研究  

Comparative Study on Predictive Modeling of Wood Chip Properties Based on Experimental and Industrial Near Infrared Spectrometer

作　　者：梁龙[1,2] 吕娜娜 吴珽 邓拥军 沈葵忠[1] 房桂干[1,4] LIANG Long;LYU Nana;WU Ting;DENG Yongjun;SHEN Kuizhong;FANG Guigan(Jiangsu Key Lab of Biomass Energy and Material,Jiangsu Co-Innovation Center of Efficient Processing and Utilization of Forest Resource,Key Lab of Chemical Engineering of Forest Products,National Forestry and Grassland Administration,National Engineering Lab for Biomass Chemical Utilization,Institute of Chemical Industry of Forest Products,CAF,Nanjing,Jiangsu Province,210042;State Key Lab of Pulp and Paper Engineering,South China University of Technology,Guangzhou,Guangdong Province,510640;Key Lab of Chemistry and Engineering of Forest Products,State Ethnic Affairs Commission,Guangxi Key Lab of Chemistry and Engineering of Forest Products,Guangxi University for Nationalities,Nanning,Guangxi Zhuang Antonomous Region,530006;Shandong Huatai Paper Co.,Ltd.,Dongying,Shandong Province,257335)

机构地区：[1]中国林业科学研究院林产化学工业研究所,江苏省生物质能源与材料重点实验室,江苏省林业资源高效加工利用协同创新中心,国家林业和草原局林产化学工程重点实验室,林木生物质低碳高效利用国家工程研究中心,江苏南京210042 [2]华南理工大学制浆造纸工程国家重点实验室,广东广州510640 [3]广西民族大学林产化学与工程国家民委重点实验室,广西林产化学与工程重点实验室,广西南宁530006 [4]山东华泰纸业股份有限公司,山东东营257335

出　　处：《中国造纸》2024年第7期7-14,共8页China Pulp & Paper

基　　金：科技创新2030-重大项目课题(2023ZD0405905);制浆造纸工程国家重点实验室开放基金(202311);林产化学与工程国家民委重点实验室暨广西林产化学与工程重点实验室开放课题重点项目(GXFK2205);泰山产业领军人才工程专项经费(tscy20200213)。

摘　　要：本研究探讨了采用工业型光谱仪和实验型光谱仪建立近红外模型用于快速预测木片材性的可行性,分析对比不同光谱仪的建模效果。针对不同材性指标分析特点,通过光谱信号校正和特征波长筛选算法优化模型预测性能,建立适宜的建模方法。结果表明,基于工业型光谱仪建立的近红外模型对木片水分、基本密度、纤维形态等物理性能指标表现出更高的预测精度,而化学组分含量的预测精度与实验型光谱仪无显著差异。工业型光谱仪木片理化材性近红外模型适用于制浆原料品质评价和工业过程分析。In this study,the feasibility of near-infrared(NIR)modeling using an industrial near infrared spectrometer for rapid predic‐tion of wood chips properties was explored and compared with the modeling results of experimental spectrometer.According to the characteristics of different wood properties,the model performance was optimized by spectral signal correction and wavelength screen‐ing algorithms,and a suitable modeling method was established.The results showed that the NIR model based on an industrial spec‐trometer exhibited higher prediction accuracy for physical proper‐ties such as moisture,basic density,and fiber length of wood chips,while the prediction accuracy of chemical component con‐tent was not significantly different from that of the experimental spectrometer.NIR model of wood chip properties based on industrial spectrometer was suitable for raw material quality evaluation and industrial process analysis.

关 键 词：近红外光谱 木片 材性分析 制浆造纸 

分 类 号：TS72[轻工技术与工程—制浆造纸工程]