基于近红外光谱分析技术的桉木纤维施胶量高低快速判别

doi:10.11707/j.1001-7488.20210814

摘要/Abstract

摘要：

目的: 基于近红外光谱分析技术，判别纤维板生产过程中施胶量高低，为施胶工艺阶段提供技术支撑。方法: 以桉木纤维为研究对象，以脲醛树脂胶黏剂施胶量高低为判别指标，对同一批桉木纤维，控制含水率5%左右，建立施胶量为无（0%）、低（3%）、中（12%）、高（20%）4种类别的近红外光谱偏最小二乘法回归模型（PLS），采用PLS-DA法判别施胶量高低，探究光谱采集方式（施胶纤维运动或静态状态）、施胶后纤维陈放时间对模型判别准确性的影响。结果: 1）随着施胶量增加，近红外光谱吸光度增大，PCA分析可区分无、低、中、高4种施胶量类别的桉木纤维；PLS-DA法能够建立相关性好、准确性高的近红外光谱模型，对未知样品的判别正确率达100%；2）静态或动态光谱采集方式不会影响模型建立，对未知样品的判别正确率达100%；施胶后纤维陈放时间对近红外光谱影响很大，施胶后长时间陈放的纤维几乎不能正确判别施胶量高低。结论: 不同施胶量桉木纤维对近红外光谱的吸光度不同，PLS-DA法可准确判别桉木纤维施胶量高低；样品处于静止或运动状态均不影响模型建立和判别的准确性，可为在线识别纤维施胶量提供一定思路。

关键词: 桉木纤维, 纤维板, 施胶量, 近红外光谱

Abstract:

Object: In order to provide technical support for applying glue working procedure and determination of resin content, this paper proposed a method by near infrared spectroscopy (NIR). Method: Eucalyptus fibers were used as the research objects and resin content was used as the discriminative index. For the same batch of eucalyptus fiber, its moisture content was about 5%. Four kinds of resin contents, including 0%, 3%, 12% and 20%, were established for four categories of none, low, medium and high level, respectively. The partial least squares regression (PLS) model between the four categories and those NIR was established. Partial least-squares discrimination analysis (PLS-DA) was used to detect the effects of collecting spectra (dynamic and static) and fibers setting time. Result: 1) With the increase of resin content, the absorbance of NIR increased. Four categories of eucalyptus fibers could be distinguished. The spectrum data were established into high relevance and accuracy models, the accuracy of the models were 100%. 2) Collecting spectra (dynamic and static) did not affect the modeling, the accuracy of the models were 100%. If fibers were set a long time, the accuracy of the models were nearly 0%. Conclusion: According to the wood fiber of different resin contents, the absorbance of NIR was different, the resin content of wood fibers could be accurately determined by PLS-DA. Through the laboratory simulation experiment, dynamic or static fibers did not affect the accuracy of model establishment and discrimination. It was expected to provide some ideas for online determination of resin content of wood fiber.

Key words: eucalyptus wood fiber, fiberboard, resin content, near infrared spectroscopy (NIR)

中图分类号:

S784

朱翰文,杜官本,杨忠,吕斌,卢胜高. 基于近红外光谱分析技术的桉木纤维施胶量高低快速判别[J]. 林业科学, 2021, 57(8): 141-146.

Hanwen Zhu,Guanben Du,Zhong Yang,Bin Lü,Shenggao Lu. Determination of Resin Content of Eucalyptus Wood Fiber by Near Infrared Spectroscopy[J]. Scientia Silvae Sinicae, 2021, 57(8): 141-146.

图/表 12

图1

图2

表1

表2

表3

表4

表5

表6

表7

图3

表8

表9

参考文献 0

	崔举庆, 全清豪, 韩书广, 等. MUF胶黏剂的稀释对降低纤维板施胶量的影响. 林业科技开发, 2015, 29 (2): 63- 65.
	Cui J Q , Quan Q H , Han S G , et al. Influence of dilution of melamine-urea-formaldehyde resin on lowering adhesive consumption in fiberboard manufacturing. China Forestry Science and Technology, 2015, 29 (2): 63- 65.
	段宇飞, 王巧华, 马美湖, 等. 基于LLE-SVR的鸡蛋新鲜度可见/近红外光谱无损检测方法. 光谱学与光谱分析, 2016, 36 (4): 981- 985.
	Duan Y F , Wang Q H , Ma M H , et al. Study on non-destructive detection method for egg freshness based on LLE-SVR and visible/near-infrared spectrum. Spectroscopy and Spectral Analysis, 2016, 36 (4): 981- 985.
	方孝荣, 章海亮, 黄凌霞, 等. 近红外光谱技术结合RCA和SPA方法检测土壤总氮研究. 光谱学与光谱分析, 2015, 35 (5): 1248- 1252. doi: 10.3964/j.issn.1000-0593(2015)05-1248-05
	Fang X R , Zhang H L , Huang L X , et al. Measurement of soil total nitrogen using near infrared spectroscopy combined with RCA and SPA. Spectroscopy and Spectral Analysis, 2015, 35 (5): 1248- 1252. doi: 10.3964/j.issn.1000-0593(2015)05-1248-05
	贾生尧, 杨祥龙, 李光, 等. 近红外光谱技术结合递归偏最小二乘算法对土壤速效磷与速效钾含量测定研究. 光谱学与光谱分析, 2015, 35 (9): 2516- 2520.
	Jia S Y , Yang X L , Li G , et al. Quantitatively determination of available phosphorus and available potassium in soil by near infrared spectroscopy combining with recursive partial least squares. Spectroscopy and Spectral Analysis, 2015, 35 (9): 2516- 2520.
	刘能文, 毛传伟, 胡帆. 我国木材市场供需情况与展望. 中国人造板, 2019, 26 (1): 33- 37.
	Liu N W , Mao C W , Hu F . Supply and demand situation and prospect of wood market in China. China Wood-Based Panels, 2019, 26 (1): 33- 37.
	刘鹏, 蒋雪松, 沈飞, 等. 近红外光谱技术的花生产毒霉菌侵染快速检测. 光谱学与光谱分析, 2017, 37 (5): 1397- 1402.
	Liu P , Jiang X S , Shen F , et al. Rapid detection of toxigenic fungal contamination in peanuts with near infrared spectroscopy technology. Spectroscopy and Spectral Analysis, 2017, 37 (5): 1397- 1402.
	孙通, 莫欣欣, 李晓珍, 等. 近红外光谱技术结合变量选择方法定性检测食用植物油中的腐霉利. 光谱学与光谱分析, 2016, 36 (12): 3915- 3919.
	Sun T , Mo X X , Li X Z , et al. Qualitative detection of procymidone in edible vegetable oils by near infrared spectroscopy and variable selection methods. Spectroscopy and Spectral Analysis, 2016, 36 (12): 3915- 3919.
	吴妍娴, 宋春风, 袁洪福, 等. 一种新型光谱多元分析模式识别方法. 光谱学与光谱分析, 2017, 37 (8): 2493- 2499.
	Wu Y X , Song C F , Yuan H F , et al. A new multivariate classification and identification method of spectroscopy. Spectroscopy and Spectral Analysis, 2017, 37 (8): 2493- 2499.
	张妍楠, 陈兰珍, 薛晓锋, 等. 基于近红外光谱检测技术鉴别洋槐蜜中掺入大米糖浆的可行性研究. 光谱学与光谱分析, 2015, 35 (9): 168- 171.
	Zhang Y N , Chen L Z , Xue X F , et al. Discrimination of rice syrup adulterant of acacia honey based using near-infrared spectroscopy. Spectroscopy and Spectral Analysis, 2015, 35 (9): 168- 171.
	张玉萍, 吕斌. 2010-2018年我国人造板产业发展简况. 中国人造板, 2019, 26 (7): 29- 30.
	Zhang Y P , Lü B . Brief introduction of China wood-based panels industry during 2010 to 2018. China Wood-Based Panels, 2019, 26 (7): 29- 30.
	Ali I , Jayaraman K , Bhattacharyya D . Effects of resin and moisture content on the properties of medium density fibreboards made from kenaf bast fibres. Industrial Crops and Products, 2014, 52, 191- 198.
	Altgen D , Bellmann M , Wascher R , et al. Enhanced urea-formaldehyde adhesive spreading on plasma treated wood particles. European Journal of Wood and Wood Products, 2016, 74 (4): 617- 620.
	Costa N , Amaral S , Alvim R , et al. Assessment of resin formulations and determination of the formaldehyde to urea molar ratio by near- and mid-infrared spectroscopy and multivariate data analysis. Journal of Applied Polymer Science, 2013, 128 (1): 498- 508.
	Henriques A , Cruz P , Ferra J M , et al. Determination of melamine content in amino resins by near-infrared spectroscopy. Wood Science and Technology, 2013, 47 (5): 939- 948.
	Henriques A , Cruz P , Martins J , et al. Determination of formaldehyde/urea molar ratio in amino resins by near-infrared spectroscopy. Journal of Applied Polymer Science, 2012, 124 (3): 2441- 2448.
	Kallakas H , Närep M , Närep A , et al. Mechanical and physical properties of industrial hemp-based insulation materials. Proceedings of the Estonian Academy of Sciences, 2018, 67 (2): 183- 192.
	Nayeri M D , Tahir P M , Jawaid M , et al. Effect of resin content and pressure on the performance properties of rubber wood-kenaf composite board panel. Fibers and Polymers, 2014, 15 (6): 1263- 1269.
	Riegler M , Gindl-Altmutter W , Hauptmann M , et al. Detection of UF resin on wood particles and in particleboards: potential of selected methods for practice-oriented offline detection. European Journal of Wood and Wood Products, 2012, 70 (6): 829- 837.
	Tsuchikawa S , Schwanninger M . A review of recent near-infrared research for wood and paper. Applied Spectroscopy Reviews, 2013, 48 (7): 560- 587.

项目Item	施胶量Resin content(%)
项目Item	0	3	12	20
施胶量 Resin content	无None	低Low	中Middle	高High
分类变量 Category variables	N[1 0 0 0]	L[0 1 0 0]	M[0 0 1 0]	H[ 0 0 0 1]

样本集 Sample		施胶量Resin content(%)
样本集 Sample		0	3	12	20
校正 Calibration	R²	0.98	0.89	0.86	0.95
	SEC	0.06	0.14	0.16	0.10
	Bias	-5.29×10^-8	1.17×10^-7	0	-7.23×10^-8
验证 Validation	R²	0.97	0.80	0.68	0.91
	SEP	0.08	0.19	0.24	0.13
	Bias	1.97×10^-3	1.62×10^-3	-7.70×10^-3	4.11×10^-3

类别 Category	预测值 Predictions(Y_P)	偏差 Deviation(D)	类别 Category	预测值 Predictions(Y_P)	偏差 Deviation(D)
N	0.98	0.08	M	0.75	0.18
N	0.88	0.10	M	0.94	0.27
N	0.96	0.08	M	0.76	0.22
N	0.99	0.10	M	1.12	0.19
N	0.99	0.11	M	1.12	0.21
N	0.99	0.10	M	0.77	0.21
N	0.93	0.08	M	0.74	0.22
N	0.98	0.08	M	0.86	0.24
N	0.78	0.13	M	0.68	0.23
N	1.03	0.09	M	0.97	0.23
L	0.75	0.13	H	0.95	0.08
L	0.74	0.17	H	0.77	0.09
L	0.73	0.17	H	0.97	0.08
L	0.98	0.15	H	1.03	0.09
L	0.81	0.21	H	0.66	0.11
L	0.89	0.17	H	0.88	0.10
L	1.00	0.20	H	1.04	0.09
L	0.63	0.17	H	1.01	0.14
L	0.66	0.23	H	0.85	0.10
L	0.92	0.17	H	0.88	0.11

样本集 Sample		施胶量Resin content(%)
样本集 Sample		0	3	12	20
校正 Calibration	R²	0.99	0.92	0.84	0.96
	SEC	0.05	0.12	0.17	0.09
	Bias	-3.40×10^-8	0	1.13×10^-7	-7.10×10^-8
验证 Validation	R²	0.98	0.85	0.73	0.95
	SEP	0.06	0.16	0.23	0.10
	Bias	-2.77×10^-4	3.29×10^-3	-3.33×10^-3	3.19×10^-4

类别 Category	预测值 Predictions(Y_P)	偏差 Deviation(D)	类别 Category	预测值 Predictions(Y_P)	偏差 Deviation(D)
N	0.95	0.06	M	0.71	0.19
N	0.95	0.07	M	0.71	0.20
N	0.90	0.07	M	0.81	0.17
N	1.03	0.08	M	0.59	0.19
N	0.93	0.06	M	0.96	0.24
N	0.97	0.07	M	1.07	0.20
N	1.03	0.07	M	0.65	0.18
N	1.04	0.04	M	0.92	0.19
N	1.02	0.08	M	0.93	0.20
N	0.98	0.07	M	1.01	0.25
L	0.75	0.21	H	0.94	0.06
L	0.98	0.19	H	0.89	0.07
L	1.06	0.17	H	0.98	0.08
L	1.01	0.13	H	1.12	0.11
L	1.06	0.17	H	1.02	0.09
L	0.78	0.16	H	0.88	0.08
L	1.11	0.18	H	0.93	0.08
L	0.95	0.13	H	0.98	0.08
L	0.91	0.21	H	0.99	0.07
L	0.95	0.19	H	1.02	0.08