基于人工蜂群算法优化SVM的NIR杉木弹性模量预测

doi:10.11707/j.1001-7488.20210117

摘要/Abstract

摘要：

目的: 利用近红外光谱分析技术，提出一种基于人工蜂群算法优化支持向量机(ABC-SVM)的木材弹性模量预测模型，为木材弹性模量无损预测提供科学参考。方法: 以294个杉木样本为试验材料，采用常规力学方法测量样本弹性模量，采集样本近红外漫反射光谱，选择350~2 500 nm光谱段，对原始数据进行15步指数平滑和一阶导数预处理，并利用主成分分析降维处理后的数据，建立偏最小二乘回归(PLS)模型、支持向量机回归(SVR)模型和人工蜂群算法优化支持向量机(ABC-SVM)模型预测杉木弹性模量，采用决定系数(R²)、均方根误差(RMSE)、平均绝对百分比误差(MAPE)和平均绝对误差(MAE)对所建模型进行对比分析。结果: PLS模型的R²为0.726 700、RMSE为6.744 9、MAPE为0.063 5、MAE为5.065 6；SVR模型的R²为0.935 305、RMSE为3.528 1、MAPE为0.023 7、MAE为1.840 9；将人工蜂群算法用于SVM参数寻优，获得的最优参数c=5.670 51、g=0.031 25，ABC-SVM模型的R²为0.935 371、RMSE为3.526 0、MAPE为0.023 7、MAE为1.840 0。3种模型均可对杉木弹性模量进行有效预测。结论: 1) 根据决定系数(R²)，SVR和ABC-SVM模型的预测性能优于PLS模型，ABC-SVM模型的预测性能最佳；2)根据均方根误差(RMSE)、平均绝对百分比误差(MAPE)和平均绝对误差(MAE)，3种模型的MAPE均在可接受范围内，ABC-SVM模型关于误差的各项指标均最小，基于ABC-SVM模型预测杉木弹性模量高效、科学。

关键词: 近红外光谱, 弹性模量, 支持向量机, 人工蜂群算法

Abstract:

Objective: Using near-infrared spectrum(NIR) analysis technology, a wood elastic modulus prediction model based on support vector machine optimized by artificial bee colony algorithm(ABC-SVM) was proposed, which was expected to provide a scientific reference for non-destructive prediction of wood elastic modulus. Method: Taking 294 samples of Cunninghamia lanceolata as experimental materials, the elastic modulus of the samples was measured by conventional mechanical method, and the near-infrared diffuse reflection spectra of the samples were collected. The 350-2 500 nm spectrum was selected and the original spectrum was preprocessed by first-order derivation and 15-step exponential smoothing. Using principal component analysis to reduce the data, then the partial least squares regression(PLS) model, support vector machine regression(SVR) model and the ABC-SVM model were established to predict the elastic modulus of Cunninghamia lanceolata. The determination coefficient(R²), the root mean square error(RMSE), the mean absolute percentage error(MAPE) and the mean absolute error(MAE) were used to evaluate the advantages and disadvantages of the three models. Result: The R² of PLS model is 0.726 700, the RMSE is 6.744 9, the MAPE is 0.063 5, and the MAE is 5.065 6. As for SVR model, the R² is 0.935 305, the RMSE is 3.528 1, the MAPE is 0.023 7, the MAE is 1.840 9. Using artificial bee colony algorithm to optimize SVM parameters, the optimal parameters are c=5.670 51, g=0.031 25. The R² of ABC-SVM model is 0.935 371, the RMSE is 3.526 0, the MAPE is 0.023 7 and the MAE is 1.840 0. All three models could effectively predict the elastic modulus of Cunninghamia lanceolata. Conclusion: 1) Observing the index of determination coefficient(R²), both SVR and ABC-SVM models have better prediction performances than PLS model, and ABC-SVM has the best prediction performance. 2)According to the obtained error results(RMSE, MAPE and MAE), the MAPE of the three models is within the acceptable range, and the ABC-SVM model has the smallest indicators for the error, indicating that the model is efficient and scientific in predicting elastic modulus of Cunninghamia lanceolata.

Key words: near-infrared spectrum(NIR), elastic modulus, support vector machine(SVM), artificial bee colony(ABC) algorithm

中图分类号:

S781.2

陈芳,程献宝,黄安民,王学顺. 基于人工蜂群算法优化SVM的NIR杉木弹性模量预测[J]. 林业科学, 2021, 57(1): 161-168.

Fang Chen,Xianbao Cheng,Anmin Huang,Xueshun Wang. Elastic Modulus Prediction of Cunninghamia lanceolata Based on Artificial Bee Colony Algorithm SVM and NIR[J]. Scientia Silvae Sinicae, 2021, 57(1): 161-168.

图/表 9

图1

图2

表1

图3

图4

图5

图6

表2

图7

参考文献 0

	郝勇, 商庆园, 饶敏, 等. 木材种类的近红外光谱和模式识别. 光谱学与光谱分析, 2019, 39 (3): 705- 710.
	Hao Y , Shang Q Y , Rao M , et al. Identification of wood species based on near infrared spectroscopy and pattern recognition method. Spectroscopy and Spectral Analysis, 2019, 39 (3): 705- 710.
	高雷阜, 佟盼. 融合改进遗传和人工蜂群的SVM参数优化算法. 计算机工程与应用, 2016a, 52 (18): 36- 39.
	Gao L F , Tong P . Algorithm of SVM parameter optimization by combining improved genetic and artificial bee colony. Computer Engineering and Applications, 2016a, 52 (18): 36- 39.
	高雷阜, 高晶, 赵世杰. 人工蜂群算法优化SVR的预测模型. 计算机工程与应用, 2016b, 52 (11): 55- 59.
	Gao L F , Gao J , Zhao S J . Forecast model of SVR optimized by artificial bee colony algorithm. Computer Engineering and Applications, 2016b, 52 (11): 55- 59.
	李璟民, 郭敏. 人工蜂群算法优化支持向量机的分类研究. 计算机工程与应用, 2015, 51 (2): 151- 155.
	Li J M , Guo M . Study on classification of artificial bee colony algorithm to optimization of support vector machine. Computer Engineering and Applications, 2015, 51 (2): 151- 155.
	刘渝根, 陈超. 基于人工蜂群算法优化支持向量机的接地网腐蚀速率预测模型. 电力自动化设备, 2019, 39 (5): 182- 186.
	Liu Y G , Chen C . Corrosion rate prediction model of grounding grid based on support vector machine optimized by artificial bee colony algorithm. Electric Power Automation Equipment, 2019, 39 (5): 182- 186.
	刘镇波, 刘一星, 于海鹏, 等. 实木板材的动态弹性模量检测. 林业科学, 2005, 41 (6): 126- 131.
	Liu Z B , Liu Y X , Yu H P , et al. Research on the dynamic modulus of elasticity measurement of lumber. Scientia Silvae Sinicae, 2005, 41 (6): 126- 131.
	毛莎莎, 曾明, 何绍兰, 等. 近红外光谱技术在水果成熟期预测中的应用. 亚热带植物科学, 2010, 39 (1): 82- 87, 96.
	Mao S S , Zeng M , He S L , et al. Application of near infrared spectra(NIRS) technology in prediction of maturity stage of fruit. Subtropical Plant Science, 2010, 39 (1): 82- 87, 96.
	谭念, 孙一丹, 王学顺, 等. 基于主成分分析和支持向量机的木材近红外光谱树种识别研究. 光谱学与光谱分析, 2017, 37 (11): 3370- 3374.
	Tan N , Sun Y D , Wang X S , et al. Research on near infrared spectrum with principal component analysis and support vector machine for timber identification. Spectroscopy and Spectral Analysis, 2017, 37 (11): 3370- 3374.
	谭念, 王学顺, 黄安民, 等. 基于灰狼算法SVM的NIR杉木密度预测. 林业科学, 2018, 54 (12): 137- 141.
	Tan N , Wang X S , Huang A M , et al. Wood density prediction of Cunninghamia lanceolata based on gray wolf algorithm SVM and NIR. Scientia Silvae Sinicae, 2018, 54 (12): 137- 141.
	王晓旭, 黄安民, 杨忠, 等. 近红外光谱用于杉木木材强度分等的研究. 光谱学与光谱分析, 2011, 31 (4): 975- 978.
	Wang X X , Huang A M , Yang Z , et al. Study on the wood grading by near infrared spectroscopy. Spectroscopy and Spectral Analysis, 2011, 31 (4): 975- 978.
	王学顺. 2010. 近红外光谱信息提取及其在木材材性分析中的应用研究. 哈尔滨: 东北林业大学博士学位论文.
	Wang X S. 2010. A study of NIR information extraction and its application in wood property analysis. Harbin: PhD thesis of Northeast Forestry University.[in Chinese]
	汪紫阳, 尹世逵, 李春旭, 等. 可见/近红外光谱技术识别树叶树种的研究. 西北林学院学报, 2019, 34 (1): 229- 236, 260.
	Wang Z Y , Yin S K , Li C X , et al. Identification of tree leaf and species by Vis/NIR spectroscopy. Journal of Northwest Forestry University, 2019, 34 (1): 229- 236, 260.
	徐峰, 刘云飞, 潘惠新, 等. 基于声-超声技术的木材弹性模量测定方法研究. 振动与冲击, 2014, 33 (4): 210- 214.
	Xu F , Liu Y F , Pan H X , et al. Wood elastic modulus determination based on acoustic-ultrasonic technique. Journal of Vibration and Shock, 2014, 33 (4): 210- 214.
	杨双艳, 杨紫刚, 张四伟, 等. 基于近红外光谱和PSO-SVM算法的烟叶自动分级方法. 贵州农业科学, 2018, 46 (12): 141- 144.
	Yang S Y , Yang Z G , Zhang S W , et al. Automatic grading method of tobacco leaves based on NIR technology and PSO-SVM algorithm. Guizhou Agricultural Sciences, 2018, 46 (12): 141- 144.
	于仕兴. 2014. 基于智能算法的支持向量机结合木材近红外光谱应用研究. 北京: 北京林业大学硕士学位论文.
	Yu S X. 2014. Support vector machine(SVM)based on intelligent algorithm combined with wood NIR application research. Beijing: MS thesis of Beijing Forestry University.[in Chinese]
	张建强, 刘维涓, 侯英. 基于稀疏表示分类和近红外光谱的烟叶自动分级研究. 光谱学与光谱分析, 2018, 38 (S1): 23- 24.
	Zhang J Q , Liu W J , Hou Y . Automatic discriminate the classes of tobacco leaves based on near-infrared spectroscopy and sparse representation classification algorithm. Spectroscopy and Spectral Analysis, 2018, 38 (S1): 23- 24.
	赵荣军, 霍小梅, 张黎. 利用近红外光谱技术预测粗皮桉木材弹性模量. 光谱学与光谱分析, 2009, 29 (9): 2392- 2395.
	Zhao R J , Huo X M , Zhang L . Estimation of modulus of elasticity of Eucalyptus pellita wood by near infrared spectroscopy. Spectroscopy and Spectral Analysis, 2009, 29 (9): 2392- 2395.
	周志茹, 赵茂程, 王正. 意杨木材弹性模量3种方法检测的比较. 福建林学院学报, 2014, 34 (4): 368- 373.
	Zhou Z R , Zhao M C , Wang Z . Comparative study of modulus of elasticity of Populus euramericana lumber with three nondestructive methods. Journal of Forest and Environment, 2014, 34 (4): 368- 373.
	Brajevic I , Tuba M . An upgraded artificial bee colony(ABC)algorithm for constrained optimization problems. Journal of Intelligent Manufacturing, 2013, 24 (4): 729- 740.
	Djemai S , Brahmi B , Bibi M O . A primal-dual method for SVM training. Neurocomputing, 2016, 211, 34- 40.
	Jones P D , Schimleck L R , Peter G F , et al. Nondestructive estimation of wood chemical composition of sections of radial wood strips by near infrared spectroscopy. Wood Science & Technology, 2006, 40 (8): 709- 720.
	Karaboga D. 2005. An idea based on honey bee swarm for numerical optimization. Technical Report-TR06, Erciyes University, Engineering Faculty, Computer Engineering Department.
	Kelley S S , Rials T G , Groom L H , et al. Use of near infrared spectroscopy to predict the mechanical properties of six softwoods. Holzforschung, 2004, 58 (3): 252- 260.
	Özparpucu M , Gierlinger N , Cesarino I , et al. Significant influence of lignin on axial elastic modulus of poplar wood at low microfibril angles under wet conditions. Journal of Experimental Botany, 2019, 70 (15): 4039- 4047.
	Unler A , Murat A , Chinnam R B . mr² PSO: a maximum relevance minimum redundancy feature selection method based on swarm intelligence for support vector machine classification. Information Sciences, 2011, 181 (20): 4625- 4641.

参数搜索范围 Parameter search range	最优的c Best c	最优的g Best g	均方根误差 RMSE	决定系数 R²(%)
2^-4~2⁴	2	0.062 50	3.679 4	92.069 0
2^-5~2⁵	4	0.031 25	3.528 1	93.530 0
2^-6~2⁶	4	0.015 62	4.253 3	91.392 4
2^-7~2⁷	11.313 7	0.007 81	4.936 8	89.296 5
2^-8~2⁸	11.313 7	0.007 81	4.936 8	89.296 5

[1]	毛学刚, 竹亮, 刘怡彤, 姚瑶, 范文义. 高空间分辨率影像与SAR数据协同特征面向对象林分类型识别[J]. 林业科学, 2019, 55(9): 92-102.
[2]	王正, 付海燕, 丁叶蔚, 曹瑜, 王韵璐, 吴晓莉, 张统越. 定向刨花板剪切模量和弹性模量动态测试[J]. 林业科学, 2019, 55(8): 136-146.
[3]	毛学刚, 邢秀丽, 李佳蕊, 谭良全, 范文义. 基于航空正射影像的面向对象林隙识别[J]. 林业科学, 2019, 55(2): 87-96.
[4]	张莹, 张晓丽, 李宏志, 黎良财. 基于改进转换分离度特征选择规则的土地覆盖分类比较[J]. 林业科学, 2018, 54(8): 88-98.
[5]	谢文, 赵小敏, 郭熙, 叶英聪, 孙小香, 匡丽花. 基于RBF组合模型的山地红壤有机质含量光谱估测[J]. 林业科学, 2018, 54(6): 16-23.
[6]	谭念, 王学顺, 黄安民, 王晨. 基于灰狼算法SVM的NIR杉木密度预测[J]. 林业科学, 2018, 54(12): 137-141.
[7]	曹正彬, 周世玉, 刘晓平, 高闯, 杜光月, 周玉成. 基于LS-SVM的采暖木质地板自然对流温度场建模[J]. 林业科学, 2018, 54(11): 29-36.
[8]	曹正彬, 刘晓平, 杜光月, 褚鑫, 刘大伟, 周玉成. 基于TWSVM和模糊的木质采暖地板蓄热温度预测模型[J]. 林业科学, 2018, 54(11): 45-52.
[9]	张帅楠, 栾启福, 姜景民. 基于无损检测技术的湿地松生长及材性性状遗传变异分析[J]. 林业科学, 2017, 53(6): 30-36.
[10]	卢万鸿, 杨桂丽, 林彦, 王楚彪, 罗建中. 桉树叶片水分状态的近红外光谱检测[J]. 林业科学, 2017, 53(5): 16-22.
[11]	毛学刚, 陈文曲, 魏晶昱, 范文义. 分割尺度对面向对象树种分类的影响及评价[J]. 林业科学, 2017, 53(12): 73-83.
[12]	毛学刚, 侯吉宇, 范文义. 林隙主被动遥感协同自动识别[J]. 林业科学, 2017, 53(11): 94-103.
[13]	王正, 顾玲玲, 高子震, 刘斌. SPF规格材弹性模量的动态测试及其概率分布[J]. 林业科学, 2015, 51(2): 105-111.
[14]	郭莹洁, 赵荣军, 钟永, 任海青. 基于早晚材的兴安落叶松成熟材力学模型 ——顺纹抗拉弹性模量[J]. 林业科学, 2014, 50(9): 118-123.
[15]	柴宇博, 孙柏玲, 刘君良, 秦特夫, 储富祥. 近红外光谱快速预测乙酰化木材的增重率[J]. 林业科学, 2014, 50(9): 124-128.