APP下载

霉变稻谷脂肪酸含量的光谱检测模型构建与优化分析

2016-04-09洪添胜李立君张仟仟中南林业科技大学机电工程学院长沙000华南农业大学工程学院南方农业机械与装备关键技术教育部重点实验室广州5062国家柑橘产业技术体系机械研究室广州5062中南林业科技大学理学院长沙000

农业工程学报 2016年1期
关键词:稻谷波段校正

文 韬,洪添胜,李立君,郭 鑫,赵 兵,张仟仟,刘 付(.中南林业科技大学机电工程学院,长沙000;2.华南农业大学工程学院南方农业机械与装备关键技术教育部重点实验室,广州5062;3.国家柑橘产业技术体系机械研究室,广州5062;.中南林业科技大学理学院,长沙000)



霉变稻谷脂肪酸含量的光谱检测模型构建与优化分析

文韬1,2,洪添胜2,3※,李立君1,郭鑫4,赵兵1,张仟仟1,刘付1
(1.中南林业科技大学机电工程学院,长沙410004;2.华南农业大学工程学院南方农业机械与装备关键技术教育部重点实验室,广州510642;3.国家柑橘产业技术体系机械研究室,广州510642;4.中南林业科技大学理学院,长沙410004)

摘要:为了实现霉变稻谷脂肪酸含量无损、快速检测,该文研究应用可见/近红外光谱技术检测霉变稻谷的脂肪酸含量。考虑到直接选用霉变稻谷可见/近红外光谱数据构建脂肪酸含量预测模型存在建模费时、预测失准等问题,研究提出了霉变稻谷脂肪酸含量的可见/近红外优化校正模型。研究中通过光谱-理化值共生距离(sample set partitioning based on joint xy distance, SPXY)算法结合偏最小二乘法初步分析了不同校正集样本预测霉变稻谷脂肪酸含量的差异;利用连续投影算法(SPA)提取了反映霉变稻谷脂肪酸含量变化的特征波段;采用偏最小二乘法(partial least square, PLS)和多元线性回归法(multivariable linear regression, MLR)分别建立了基于特征波段光谱反射值的霉变稻谷脂肪酸含量预测模型,并对比分析了采用SPXY样本集划分的模型预测效果。结果表明:采用SPXY法筛选出的65个校正集样本分布与初始校正集相近,脂肪酸含量变化范围为18.55~127.26 mg,其标准差为32.39;SPA算法最终从256个全光谱波段中优选出9个特征波段,实现了光谱数据的压缩;分别建立的SPXY-SPA-PLSR模型和SPXY-SPA-MLR模型预测霉变稻谷脂肪酸含量相关系数RP为0.922 1和0.915 9,预测均方根误差RMSEP为13.889 3和14.261 0;SPXY筛选校正集所构建模型预测精度与初始校正集所建模型相当,但校正集样本数量减少为初始校正集的41%,运算时长缩短为初始样本集的32%,提高了模型的校正速度。

关键词:模型;光谱检测;农业;霉变稻谷;脂肪酸;可见/近红外光谱;特征波段;样本集划分

文韬,洪添胜,李立君,郭鑫,赵兵,张仟仟,刘付.霉变稻谷脂肪酸含量的光谱检测模型构建与优化分析[J].农业工程学报,2016,32(01):193-199.doi:10.11975/j.issn.1002-6819.2016.01.027 http://www.tcsae.org

Wen Tao, Hong Tiansheng, Li Lijun, Guo Xin, Zhao Bing, Zhang Qianqian, Liu Fu.Optimization analysis and establishment of spectra detection model of fatty acid contents for mould paddies[J].Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2016, 32(01): 193-199.(in Chinese with English abstract)doi:10.11975/j.issn.1002-6819.2016.01.027 http://www.tcsae.org

中国农业工程学会高级会员:文韬(E041200816S)

中国农业工程学会高级会员:洪添胜(E041200036S)

0 引言

稻谷霉变过程实质上就是微生物以稻谷为营养基质,进行消化、吸收和利用的物质代谢和能量代谢的生物化学反应,其中脂肪酸是一种比较稳定的代谢产物,容易在霉变的稻谷中积累,从而导致稻谷中脂肪酸值增高[1-3]。因此,脂肪酸含量的变化可以较好地表征稻谷霉变的程度。

现有的脂肪酸值测定主要采取传统的化学分析方法,该方法在分析稻谷脂肪酸含量时需要添加化学试剂对稻谷本身实施破坏性检测,处理反应周期较长,易造成对环境的污染,难以达到快速检测的要求[4-6]。因此,研究一种快速、无损检测稻谷中脂肪酸含量的方法对于诊断稻谷霉变具有重要现实意义。可见光/近红外光谱(Vis/ NIR)分析技术可以通过物质内部成分对可见/近红外光的特征吸收实现定性和定量分析,是近年来发展起来的一种高效快速的现代分析技术。目前,国内外研究学者应用其对农作物生长及环境监测和农产品品质检测开展了大量的研究工作[7-15]。近年来,有学者将该技术应用于稻谷内部成分测定研究并取得了初步进展。陆艳婷等[16]应用近红外光谱法建立了粳稻直链淀粉含量预测模型,模型决定系数为0.813,预测均方根误差为2.74。郭咏梅等[17]采用偏最小二乘法建立糙米蛋白质预测校正模型,模型决定系数为0.899。张强等[18]应用近红外光谱结合化学分析方法测定储藏稻谷中黄曲霉毒素B1含量,建立稻谷黄曲霉毒素B1支持向量机模型,模型决定系数达0.913。然而,上述研究在采用光谱数据建模时,为了使所建模型稳定准确,往往要求校正集中样本数量较多,但未深入研究校正集样本的质量,可能导致大量的数据样本间存在差异过小或相同的状况,耗费了大量的建模时间。

本研究选取稻谷霉变生物化学反应产生的代谢产物脂肪酸为研究对象,通过分时段测定不同霉变时期稻谷的光谱信息和相应的脂肪酸值,采用光谱-理化值共生距离(sample set partitioning based on joint x-y distance, SPXY)算法对稻谷可见/近红外光谱初始校正集进行划分,筛选出具有代表性的校正样本集,并选取该校正集建立基于特征波段光谱反射值的稻谷脂肪酸含量预测模型,通过与未经筛选初始校正集建模结果比较,验证稻谷脂肪酸含量可见/近红外光谱校正模型的优化效果。

1 材料与方法

1.1稻谷样本制备及脂肪酸测定

本研究所选用的稻谷样本为C两优34156晚稻,含水率为14.2%,由湖南农业大学提供。考虑到稻谷霉变程度对其脂肪酸含量的动态影响,人工选取完整、无霉变、未发芽的稻谷样本放置于恒温恒湿箱,依据稻谷霉菌适宜滋生的条件,设定恒温恒湿箱的温度30℃,相对湿度90%,进行霉变样本培育[19]。上述霉变稻谷的制备过程按照稻谷储藏过程中理化特性与感官指标随时间的变异情况,人工将培育阶段划分成3个周期,每个时间周期约为10 d[20],共得到不同霉变程度的稻谷样本各50份。制备的样本按照GB/T 20569-1995《谷物制品脂肪酸值测定法》测定稻谷脂肪酸含量[21],并将其作为建模的标准参考值。

1.2光谱信息校正及数据采集

考虑到采集的稻谷谷粒较小且样本数量较多,本研究采用Hypersis农产品高光谱仪(Hypersis-VNIR-PFH,卓立汉光,北京)完成稻谷光谱信息采集。该仪器主要包括图像采集卡和高性能光谱相机(V10E-QE)、配套光源、PSA300-X型电动位移台、集成驱动控制台(高速IMS步进电机)及暗箱等部件。光谱仪设置理想曝光时间20 ms,移动平台运行速度14.6 mm/s,扫描距离150 mm,光谱范围380~1 000 nm,光谱分辨率2.8nm。由于光源的强度分布不均及暗电流噪声存在,每次采样均需利用全黑、全白标定图像对扫描的稻谷图像进行校正,校正公式如(1)所示:

式中Ia为校正后的稻谷扫描图像;Io为校正前的稻谷扫描图像;Iw为全白标定图像;Ib为全黑标定图像。光谱数据采集实验中,将稻谷样本平铺固定于反射率接近于0的黑色底板上(如图1所示),黑色底板置于在载物台,在电机的驱动下,样本垂直于镜头纵向移动。高光谱相机同时获得样本在各波长处的光谱信息和图像信息,每粒稻谷采集得到256个波段的图像。

图1 稻谷样本在光谱检测载物台上分布Fig.1 Placement for paddy samples on workbench of spectral detection

利用遥感图像处理平台(environment for visualizing images,ENVI)选取矩形载物台上的稻谷作为感兴趣区域(region of interest, ROI),稻谷脂肪酸含量的实测值均与所选的ROI区域对应。通过计算ROI的各个像素点的光谱响应平均值来估算稻谷脂肪酸的相对反射率。

2 数据处理与模型建立

2.1SPXY样本划分方法

SPXY算法是一种基于统计基础的样本集选择方法,能使校正集最大程度地表征样本均匀分布,以提高模型稳定性,试验证明SPXY法能有效地用于近红外光谱模型的建立[22]。该算法的计算过程为:使用近红外光谱-理化值的共生距离作为划分依据,SPXY在样本间欧式距离计算时将x变量(样本近红外光谱值)和y变量(样本理化值)同时考虑在内,其x和y距离公式如(2)、(3)所示:

式中xp(j)为样本p实测的相对光谱反射值;xq(j)为样本q实测的相对光谱反射值;j为对应的近红外光谱波长,nm;yp为样本p实测的脂肪酸含量,mg/100 g;yq为样本q实测的脂肪酸含量,mg/100 g。

SPXY算法逐步选择的过程中用dxy(p,q)代替了dx(p,q),同时为了确保样本在x和y空间中具有相同的权重,将dx(p,q)和dy(p,q)分别除以他们在数据集中的最大值,其标准化的xy距离公式如(4)所示:

2.2校正模型建立与评价

本研究运用光谱技术建立近红外光谱与稻谷脂肪酸含量之间的校正模型与评价过程如图2所示:1)在稻谷样本制备的正常、霉变初期、中期和后期4个区段随机选取共计45个样本作为模型预测集;2)对模型初始校正集和预测集样本的光谱数据进行savitzky-golay(SG)平滑[23]预处理,减弱噪声影响;3)采用SPXY算法从剩余155个样本中筛选出具有差异性及代表性的校正集样本用于模型建立;4)采用连续投影算法(successive projections algorithm, SPA)[24]提取光谱数据的特征波段,消除原始光谱矩阵中冗余的信息,实现光谱数据压缩;5)采用偏最小二乘法(partial least square, PLS)[25]和多元线性回归法(multivariable linear regression, MLR)[26],分别建立光谱特征波段与脂肪酸含量之间的校正模型;6)使用预测集数据对校正模型进行检验和评价。模型建立过程中,使用相关系数Rp和预测均方根误差(RMSEP)等指标来评价模型质量,其相应的计算公式如(5)和(6)。

式中n为样本数;yTi为样本实测值;ypi为样本预测值;y-a为样本实测平均值。

图2 校正模型建立与评价流程Fig.2 Calibration model establishment and evaluation process

3 结果与分析

3.1制备样本脂肪酸含量统计

本研究制备的200个稻谷样本分布于正常期、霉变初期、中期和后期4个阶段。通过理化试验,测得不同时期稻谷脂肪酸含量分布如图3所示,4个不同时期的霉变稻谷脂肪酸含量具有不同梯度分布,脂肪酸含量在稻谷霉变的初期和中期上升速率较快,到达后期上升速率基本趋于平缓,符合文献提出的研究结论[27]。上述结果说明试验制备的样本具有一定代表性,脂肪酸含量可作为检测稻谷发生霉变的依据。

图3 不同霉变时期稻谷脂肪酸含量分布图Fig.3 Distribution map of fatty acid value in paddy for different mould stage

3.2SPXY算法划分校正集比较

校正集样本的选取与确定直接影响模型的预测精度。本研究采用SPXY算法对稻谷初始校正集样本进行筛选,指定样本数N范围选为35~155,步长为10,分别试建全光谱波段的PLS模型,根据模型预测集的相关系数Rp和预测均方根误差(RMSEP)值,确定最佳的校正集样本数量。研究获得的Rp和RMSEP随校正集样本数量变化曲线如图4所示。

从图中曲线的变化趋势可知,Rp曲线随着校正集样本数N的增加呈递增趋势,与之相对应的RMSEP曲线呈递减趋势,N的取值在35~65范围内,Rp变化差异明显,N取值在65~155范围内,Rp变化趋于平缓,N取值65为该变化曲线的数据拐点,相对应的(Rp,RMSEP)为(0.928 6,13.085 6),当N取值为135时,Rp达到极大值0.943 9,RMSEP降为极小值11.849 2。综合上述Rp和RMSEP曲线的变化趋势可知,校正集样本数N取值为65以后,预测值与真实值之间的相关系数和均方根误差均无明显差异,考虑到建模的运算量,本研究最终通过SPXY算法筛选出65个样本组成模型校正集。

图4 Rp和RMSEP随校正集样本数量变化曲线Fig.4 Variation curve of Rp and RMSEP values followed by calibration sample numbers change

预测集、初始校正集和SPXY筛选校正集的样本脂肪酸含量统计结果如表1所示。由表中统计结果可知经SPXY法筛选出的校正集样本脂肪酸含量变化范围与未经挑选的初始校正集相同,并且标准差为32.39,与初始校正集相近,说明SPXY法筛选后的校正集样本具有一定代表性。

表1 不同集稻谷肪酸含量分布Tab.1 Fatty acid values distribution in different sample set

3.3校正模型特征波段选取

在利用Vis/NIR光谱建模过程中,通过特定方法筛选特征波段或波段区间有可能得到更好的定量校正模型[28]。本研究利用连续投影算法(SPA)对稻谷的校正模型进行光谱特征波段选取,指定波段数N范围为2~24[29],根据校正集的内部交叉验证均方根误差RMSECV值确定最佳的光谱特征波段个数。

稻谷校正集样本的原始光谱经过SG数据平滑,从256个波段中共优选出9个特征波段,分别是392、404、430、442、619、636、870、885和899 nm,如图5所示。

上述研究结果显示,利用SPA算法选择光谱波段实现了光谱数据的压缩,降低了模型的复杂度。

图5 稻谷平滑光谱模型特征波段数确定和优选Fig.5 Optimal selection of characteristic wavelengths and numbers for paddy SG smoothing model

3.4预测模型建立与结果分析

1.2.1.2 成立在职培训管理组:由区护士长任组长,指定护理组长一对一带教;根据CSSD制定的岗前培训计划,专人负责,落实到位。

采用SPXY算法选取的65份稻谷校正集样本,经过SG平滑对全波段光谱进行预处理后,将SPA算法优选的特征波段下的光谱反射率作为PLSR模型和MLR模型的输入,利用预测集的45份稻谷样本检验构建模型的预测效果,其模型预测值与实测值之间离散分布如图6所示。

图6 不同模型预测值与实测值相关性Fig.6 Correlation analysis between predicted value and actual value for different models

由图6可知,利用SPA算法选取的特征波段分别建立的SPXY-SPA-PLSR模型和SPXY-SPA-MLR模型预测精度RP分别为0.922 1和0.915 9,模型预测均方根误差RMSEP分别为13.889 3和14.261 0,说明SPA算法所优选出的波段是能够基本表征待测组分信息,模型对不同霉变时期的稻谷脂肪酸含量均具有较强的预测能力。为了进一步说明采用SPXY算法筛选校正集对稻谷脂肪酸含量近红外校正模型的影响,在相同的条件,本研究将SPXY算法筛选的校正集所建的SPA-PLSR模型与初始校正集所建的SPA-PLSR模型进行了比较,相应的比较结果如表2所示。

表2 不同校正集预测的稻谷脂肪酸含量比较Tab.2 Comparison of predicted fatty acid values in different calibration models

表2的对照结果表明,选用SPXY筛选校正集,经SG平滑光谱预处理后建立的SPA-PLSR模型,校正时其内部验证的相关系数RC和均方根误差RMSEC分别为0.915 1、12.957 3;其外部验证的相关系数RP和均方根误差RMSEP分别为0.922 1、13.889 3;SPXY筛选校正集所构建模型预测精度与初始校正集所建模型相当,但校正集样本数量减少为初始校正集的41%,运算时长缩短为初始样本集的32%,说明经过SPXY算法筛选后的校正集样本是基本能够正确反映初始样本集信息,较好的消除冗余样本,提高了模型的校正速度。

4 结论

本文研究了霉变稻谷脂肪酸含量的可见/近红外光谱校正模型构建和优化方法,并通过试验进行了验证,得到以下结论:

1)试验制备的4个不同霉变时期的稻谷脂肪酸含量具有不同梯度分布。脂肪酸含量在稻谷霉变的初期和中期上升速率较快,到达后期基本趋于平缓。脂肪酸含量可作为检测稻谷霉变的依据。

2)采用SPXY法筛选出的65个校正集样本脂肪酸含量变化范围与未经挑选的初始155个校正集相同,并且标准差为32.39,与初始校正集相近,说明SPXY法筛选后的校正集样本具有一定代表性。

3)经SG平滑处理后的光谱数据,利用SPA算法进行光谱特征波段选择,最终从256个波段中优选出9个光谱波段,极小化光谱变量之间的共线性影响,实现了光谱数据的压缩,降低了模型的复杂度。

4)分别建立的SPXY-SPA-PLSR模型和SPXY-SPAMLR模型预测霉变稻谷脂肪酸含量RP为0.922 1和0.915 9,预测均方根误差RMSEP为13.889 3和14.261 0,说明模型对不同霉变时期的稻谷脂肪酸含量均具有较强的预测能力;SPXY筛选校正集所构建的SPA-PLSR模型预测精度与初始校正集所建的SPA-PLSR模型相当,但校正集样本数量减少为初始校正集的41%,运算时长缩短为初始样本集的32%,进一步说明经过SPXY算法筛选后的校正集样本是能够正确反映初始样本集信息,较好的消除冗余样本,提高了模型的校正速度。

[参考文献]

[1] Aibara S, Ismail I A.Changes in rice bran lipids and fatty acids during storage[J].Agriculture Biology Chemistry,1986, 50(3): 665-673.

[2]蔡静平,等.粮油微生物学[M].中国轻工业出版社,2002.

[3]成岩萍.浅析粮食微生物与粮食储藏的关系[J].粮油食品科技,2005, 13(1):28-29.Cheng Yanping.Ontherelationshipbetween grain microorganism and grain storage[J].Science and technology of cereals, oils and foods, 2005, 13(1): 28-29.(in Chinese with English abstract)

[4]包清彬,猪谷富雄.储藏条件对糙米理化特性影响的研究[J].农业工程学报,2003, 19(6):25-27.Bao Qinbing, Tomio ITANI.Influence of storage conditions on physicochemical characteristic of brown rice[J].Transactions of the Chinese Society of Agricultural Engineering(Transactions of the CSAE), 2003, 19(6): 25-27.(in Chinese with English abstract)

[5]邵龙,邵怀宗,彭启琮.稻谷脂肪自动酸滴定仪设计与实现[J].电子测量与仪器学报,2008, 22(6): 97-102.Shao Long, Shao Huaizong, Peng Qicong.Design and realization of rice fatty acid automatic titrator[J].Journal of electronic measurement and instrument, 2008, 22(6): 97-102.(in Chinese with English abstract)

[6]周显青,张玉荣.储藏稻谷品质指标的变化及其差异性[J].农业工程学报,2008, 24(12): 238-242.Zhou Xianqing, Zhang Yurong.Changes and differential analysis of the quality indexes of stored paddy[J].Transactions of the Chinese Society of Agricultural Engineering(Transactions of the CSAE), 2008, 24(12):238-242.(in Chinese with English abstract)

[7]宋海燕,岑海燕,应霞芳,等.基于光谱技术的禽畜污水化学需氧量快速测定方法的研究[J].农业工程学报,2006, 22(6): 148-151.Song Haiyan, Cen Haiyan, Ying Xiafang, et al.Approach to rapid detection of chemical oxygen demand in livestock waste water based on spectroscopy technology[J].Transactions of the Chinese Society of Agricultural Engineering(Transactions of the CSAE), 2006, 22(6): 148-151.(in Chinese with English abstract)

[8]洪添胜,乔军,Ning Wang,等.基于高光谱图像技术的雪花梨品质无损检测[J].农业工程学报,2007, 23(2):151-155.Hong Tiansheng, Qiao Jun, Ning Wang, et al.Non-destructive inspection of Chinese pear quality based on hyperspectral imaging technique[J].Transactions of the Chinese Society of Agricultural Engineering(Transactions of the CSAE), 2007, 23 (2): 151-155.(in Chinese with English abstract)

[9] He Yong, Huang Min, Song Haiyan, et al.Prediction of soil macronutrients content using near-infrared spectroscopy [J].Computers and Electronics in Agriculture, 2007, 58(2): 144-15.

[10]李晓丽,唐月明,何勇,等.基于可见/近红外光谱的水稻品种快速鉴别研究[J].光谱学与光谱分析,2008, 28(3): 578-581.Li Xiaoli, Tang Yueming, He Yong, et al.Discrimination of varieties of paddy based on Vis/NIR spectroscopy combined with chemometrics[J].Spectroscopy and spectral analysis, 2008, 28 (3): 578-581.(in Chinese with English abstract)

[11] Liu Fei, Zhang Fan, Jin Zonglai, et al.Determination of acetolactate synthase activity and protein content of oilseed rape (Brassica napus L.)leaves using visible/near-infrared spectroscopy[J].Analytica Chimica Acta, 2008, 629(1-2): 56-65

[12]周子立,张瑜,何勇,等.基于近红外光谱技术的大米品种快速鉴别方法[J].农业工程学报,2009, 25(8): 131-135.Zhou Zili, Zhang Yu, He Yong, et al.Method for rapid discrimination of varieties of rice using visible NIR spectroscopy [J].Transactions of the Chinese Society of Agricultural Engineering(Transactions of the CSAE), 2009, 25(8): 131-135.(in Chinese with English abstract)

[13] Dissing B S, Nielsen M E, Ersboll B K, et al.Multispectral imaging for determination of astaxnthin concentration in salmonids[J].PLoS One, 2011, 6(5):19-32.

[14]梅慧兰,邓小玲,洪添胜,等.柑橘黄龙病高光谱早期鉴别及病情分级[J].农业工程学报,2014, 30(9):140-147.Mei Huilan, Deng Xiaoling, Hong Tiansheng, et al.Early detection and grading of citrus huanglongbing usinghyperspectral imaging technique[J].Transactions of the Chinese Society of Agricultural Engineering(Transactions of the CSAE), 2014, 30(9): 140-147.(in Chinese with English abstract)

[15]邹小波,张小磊,石吉勇,等.基于高光谱图像的黄瓜叶片叶绿素含量分布检测[J].农业工程学报,2014,30(13):169-175.Zou Xiaobo, Zhang Xiaolei, Shi Jiyong, et al.Detection of chlorophyll content distribution in cucumber leaves based on hyperspectral imaging[J].Transactions of the Chinese Society of Agricultural Engineering Transactions of the CSAE, 2014, 30 (13):169-175.(in Chinese with English abstract)

[16]陆艳婷,金庆生,叶胜海,等.应用近红外光谱技术快速测定粳稻品种的直链淀粉含量[J].中国粮油学报,2007,22(3):149-153.Lu Yanting, Jin Qingsheng, Ye Shenghai, et al.Application of near infrared spectroscopy for rapid analysis of amylose content in short-grain rice[J].Journal of the Chinese Cereals and Oils Association, 2007, 22(3):149-153.(in Chinese with English abstract)

[17]郭咏梅,李华慧,李少明,等.糙米蛋白质含量与矿质元素含量的相关分析及NIRS模型的建立[J].植物遗传资源学报,2013,14(1):173-178.Guo Yongmei, Li Huahui, Li Shaoming, et al.Correlation analysis of protein content and mineral content in brown rice and establishment of the math model for the NIRS analysis [J].Journal of Plant Genetic Resources, 2013, 14(1): 173-178.(in Chinese with English abstract)

[18]张强,刘成海,孙井坤,等.基于支持向量机稻谷黄曲霉毒素B1近红外无损检测[J].东北农业大学学报,2015,46(5):84-88.Zhang Qiang, Liu Chenghai, Sun Jingkun, et al.Near-infrared spectroscopy nondestructive determination of aflatoxin B1 in paddy rice based on support vector machine regression [J].Journal of Northeast Agricultural University, 2015, 46(5): 84-88.(in Chinese with English abstract)

[19]周建新.论粮食霉变中的生物化学[J].粮食储藏,2004, 32(1): 9-12.Zhou Jianxin.The biochemistry during grain mildewing[J].Grain storage, 2004, 32(1): 9-12.(in Chinese with English abstract)

[20]张瑛,吴先山,吴敬德,等.稻谷储藏过程中理化特性变化的研究[J].中国粮油学报,2003, 29(5): 565-566.Zhang Ying, Wu Xianshan, Wu Jingde, et al.Study on physical and chemical characters in rice storage[J].Journal of the Chinese cereals and oils association, 2003, 18(6):565-566.(in Chinese with English abstract)

[21] GB/T 15684,《谷物制品脂肪酸值测定法》测定稻谷脂肪酸含量[S].

[22]展晓日,朱向荣,史新元,等.SPXY样本划分法及蒙特卡罗交叉验证结合近红外光谱用于橘叶中橙皮苷的含量测定[J].光谱学与光谱分析,2009,29(4):964-968.Zhan Xiaori, Zhu Xiangrong, Shi Xinyuan, et al.Spectroscopy with SPXY algorithm for sample subset partitioning and mote carlo cross validation [J].Spectroscopy and spectral analysis, 2009, 29(4): 964-968.(in Chinese with English abstract)

[23] Bauriegel E, Giebel A, Geyer M, et al.Early detection of Fusarium infection in wheat using hyper-spectral imaging [J].Computers and Electronics in Agriculture, 2011, 75(2): 304-312.

[24] Moreira E D T, Pontes M J C, Galváo R K H, et al.Nearinfrared reflectance spectrometry classification of cigarettes using the successive projections algorithm for variable selection [J].Talanta, 2009, 79: 1260-1264.

[25] Fei Liu, Fan Zhang, Zonglai Jin, et al.Determination of acetolactate synthase activity and protein content of oilseed rape (Brassicanapus L.)leavesusingvisible/near-infraredspectroscopy [J].Analytica chimica acta, 2008(629): 56-65.

[26]刘天玲,苏琪雅,孙群,等.基于NIR分析和模式识别技术的玉米种子识别系统[J].光谱学与光谱分析,2012,32(6):1550-1553.Liu Tianling, Su Qiya, Sun Qun, et al.Recognition of corn seeds based on pattern recognition and near infrared spectroscopy technology[J].Spectroscopy and spectral analysis,2012, 32(6):1550-1553.(in Chinese with English abstract)

[27]杨晓蓉,周建新,姚明兰,等.不同储藏条件下稻谷脂肪酸值变化和霉变相关性研究[J].粮食储藏,2006,35(5):49-52.Yang Xiaorong, Zhou Jianxin, Yao Minglan, et al.Study on a correlation between the fatty acid value change and mould of paddy in different storage conditions[J].Grain Storage, 2006, 35 (5): 49-52.(in Chinese with English abstract)

[28]褚小立,袁洪福,陆婉珍.近红外分析中光谱预处理及波长选择方法进展与应用[J].化学进展,2004,16(4):528-542.Chu Xiaoli, Yuan Hongfu, Lu Wanzhen.Progress and appliaction of spectral data pretreatment and wavelength selection methods in NIR analytical technique[J].Progress in chemistry, 2004, 16 (4): 528-542.(in Chinese with English abstract)

[29]高洪智,卢启鹏,丁海泉,等.基于连续投影算法的土壤总氮近红外特征波长的选取[J].光谱学与光谱分析,2009,29 (11):2951-2954.Gao Hongzhi, Lu Qipeng, Ding Haiquan, et al.Choice of characteristic near-infrared wavelengths for soil total nitrogen based on successive projection algorithm[J].Spectroscopy and spectral analysis, 2009, 29(11): 2951-2954.(in Chinese with English abstract)

·农业生物环境与能源工程·

Optimization analysis and establishment of spectra detection model of fatty acid contents for mould paddies

Wen Tao1,2, Hong Tiansheng2,3※, Li Lijun1, Guo Xin4, Zhao Bing1, Zhang Qianqian1, Liu Fu1
(1.School of Mechanical and Electrical Engineering, Central South University of Forestry and Technology, Changsha 410004, China; 2.Key Laboratory of Key Technology for South Agricultural Machinery and Equipment, Ministry of Education, Engineering College of South China agricultural University, Guangzhou 510642, China; 3.Division of Citrus Machinery, China Agriculture Research System,Guangzhou 510642, China;4.School of Science, Central South University of Forestry and Technology, Changsha 410004, China)

Abstract:Fatty acids were stable metabolites and easily accumulated in paddies mould process which could better express mould extend of paddies.To achieve the non-destructive and rapid detection in fatty acid contents(FAC)for mould paddies, the detection of FAC for mould paddies was studied using the Visible/Near-infrared reflectance(Vis/NIR)spectral technology.The variety C liang-you 34156 late rice was used as paddy samples, which was obtained from Hunan Agricultural University.The mould paddy cultivating test and FAC determination experiments were carried out from October 15, 2014 to January 31, 2015 in Central South University of Forestry and Technology.Normal and complete paddies were selected and loaded into 200 sample boxes by numbers.Each sample box was loaded with 100g weights.Among them, 50 sample boxes were put into the No.A constant temperature humidity chamber to store according to requirements of cereal storage(temperatures:10℃, humidities:15%)and the remaining 150 sample boxes were placed in the No.B constant temperature humidity chamber to cultivate according to mould paddies breeding conditions(temperatures: 30℃, humidities: 90%).In view of the FAC variations affected by degree of mould paddies, the cultivated process of mould paddies was divided into three periods for better representative and generalization of samples.It was 10 days in each period, and 50 pieces of mould paddy samples in different degrees were measured during the preparation process.The Vis/NIR-infrared spectral detection testing for mould paddy samples were performed in corresponding periods in South China Agricultural University.A Vis/NIR-infrared spectral device for agro-products was used for scanning of reflectance spectra for paddies.Taking into consideration that the disadvantage of time consumption and low precision in building the model, the Vis/NIR calibration model of the fatty acid in mould paddies was proposed using sample set partition based on joint X-Y distances(SPXY)algorithm in sample set.The difference of predicting FAC in mould paddies from different calibration set was preliminarily analyzed using the combination of the SPXY algorithm and the partial least-squares regression(PLSR)algorithm.The successive projection algorithm(SPA)was applied to obtain the characteristic wavelength which indentified the variation of FAC in mould paddies.The predicted models of the FAC in mould paddies based on reflection values of characteristic wavelengths were built using the PLSR and multiple linear regression(MLR)methods, and then the prediction performance were compared between the model built by the selected calibration sample set and the model built by initial calibration sample set.The results indicated that FAC of paddies which were determined from different stages had a varying gradient distribution.The related FAC from the normal stage, early stage of mould, middle stage of mould and last stage of mould ranged from 18.55 to 24.40 mg, from 27.03 to 80.90 mg, from 84.44 to 127.26 mg, and from 101.09 to 124.88 mg, respectively.The range of FAC in 65 calibration sample sets by the SPXY was consistent with in 155 initial calibration sample sets.The standard deviation of FAC in 65 calibration sample sets was 32.39, which was close to the initial calibration sample sets.Nine characteristic wavelengths were selected from 256 full wavelengths by the SPA, which fulfilled the spectral data compression.The prediction set correlation coefficient(Rp)of the SPXY-SPA-PLSR model and the SPXYSPA-MLR model were 0.922 1 and 0.915 9 and their prediction mean square root errors were 13.889 3 and 14.261 0, respectively.The model prediction precision built by the SPXY calibration set was close to its by the initial calibration, while the number of the SPXY calibration set was dropped to 41% and its computing time was reduced to 32% compared with the initial calibration, which may contribute to speed up the model establishment.

Keywords:models; spectrometry; agriculture; mould paddies; fatty acids; Vis/NIR spectra; characteristic wavelengths; sample set selection

通信作者:※洪添胜(1955-),男,广东梅县人,博士,教授,博士生导师,主要从事农业工程、机电一体化和信息技术应用研究。广州华南农业大学工程学院510642。Email:tshong@scau.edu.cn

作者简介:文韬(1983-),男,湖南长沙人,博士,副教授,主要从事农业工程、机电一体化和信息技术应用研究。长沙中南林业科技大学机电工程学院410004。Email:wt207@sina.com

基金项目:国家自然科学基金(31401281);湖南省自然科学基金(14JJ3115);湖南省大学生研究性学习和创新性实验计划项目(湘教通[2014]248号);湖南省高校科技创新团队支持计划(2014207)

收稿日期:2015-08-19

修订日期:2015-11-13

中图分类号:S123;S51

文献标志码:A

文章编号:1002-6819(2016)-01-0193-07

doi:10.11975/j.issn.1002-6819.2016.01.027

猜你喜欢

稻谷波段校正
谦卑的稻谷
玉米价疯涨 稻谷也凑热闹
劉光第《南旋記》校正
基于PLL的Ku波段频率源设计与测试
在Lightroom中校正镜头与透视畸变
小型化Ka波段65W脉冲功放模块
机内校正
田野稻谷香
春节过后 稻谷行情稳中趋弱
M87的多波段辐射过程及其能谱拟合