基于多源信息特征的泰国香米快速鉴别

泰国香米品质优良,具有较高的食味品质和营养价值,深受各国消费者喜爱,然而,市场上时常出现以次充好、掺伪等不良现象。建立泰国香米的快速鉴定方法,对于促进优质大米产业的可持续健康发展具有重要的意义。本研究基于泰国香米的近红外光谱和常规指标融合的多源信息,将155个泰国香米和194个非泰国香米样品一阶导数预处理后的近红外光谱与常规指标特征向量融合构成349行、14列向量矩阵,作为支持向量机分类器的输入向量矩阵,建立多源信息的融合模型,校正集和验证集模型的识别率均是100%,解决了近红外光谱等常用方法鉴别泰国香米时存在的准确率偏低的问题,实现了对泰国香米快速、准确地定性鉴定,在其掺伪识别方面具有重要的应用价值。     

基于BP神经网络的烟草叶片质体色素高光谱反演

为促进高光谱遥感技术在烟叶品质监测中的应用,以云烟87为研究对象,设置不同的光质处理,测定了不同生育时期烤烟叶片的质体色素(叶绿素a、叶绿素b和类胡萝卜素)含量(质量分数)及相应的叶片光谱反射率。采用小波能量系数提取法将原始光谱数据从2 151个降维为21个,并分别建立了以质体色素实测值作为输出因子,降维后的光谱数据作为输入因子的BP神经网络预测模型,训练函数采用L-M优化算法函数trainlm,输入层和输出层传递函数分别为S型正切传递函数tansig和线性传递函数purelin,叶绿素a、叶绿素b和类胡萝卜素的神经网络的隐含层节点数分别为27、32和45。结果表明,叶绿素a、叶绿素b和类胡萝卜素各模型决定系数R2分别为0.84、0.86和0.76;均方根误差RMSE分别为0.12、0.14和0.10;相对误差绝对平均值K分别为0.23、0.21和0.15。3种质体色素模型的拟合精确度均较高,误差较小,整体效果良好。     

泰国茉莉香米的检测方法

为了保证泰国茉莉香米品种真实性和纯度,需要建立相应的检测方法并制定方法标准。对新颁布的泰国茉莉香米品种鉴定和纯度检测行业标准所涉及的技术背景和技术要素进行阐述,以促进该方法在相关实验室的应用,为泰国茉莉香米的进口和市场管理提供技术保障。     

基于新误差函数的豆油脂光谱BP模式识别研究

通过构造新的误差函数,实现了基于近红外透射光谱结合BP神经网络辨识合格豆油和地沟油的方法。在10 000 cm-1~3 500 cm-1范围内采集优质豆油和地沟油透射光谱,对光谱数据依次作出卷积平滑、基线校正预处理,采用主成分分析方法提取8个主成分(累计贡献率达到99.62%)作为神经网络输入神经元,建立三层8个输入,1个输出,8个隐节点的BP神经网络模型,模型能够有效辨识豆油脂和地沟油,类别预测正确率为100%,其中豆油脂相对偏差在0.4%以内,地沟油相对偏差控制在1.0%以内。新的误差函数可以有效避免BP神经网络陷入局部极小值而未能收敛,并且提高了网络收敛速度。     

基于人工神经网络的燃油烘丝机出口水分预测

采用BP人工神经网络模型,以燃油烘丝机控制参数为输入变量,以燃油烘丝机出口水分为输出变量,建立神经网络模型,模拟输出变量的非线性关系。结果表明:在样本空间内,建立的6个隐含层神经元的BP神经网络模型对燃油烘丝机出口水分具有较好的预测能力,且精度较高,预测值与实际值平均偏差0.0074,可用于燃油烘丝机出口水分的计算。     

基于BP神经网络的非织造材料基复合吸声体吸声系数预测

以涤纶针刺非织造材料和聚丙烯熔喷非织造材料为研究对象,通过实验获得其物理结构参数,并将复合前后非织造材料厚度、面密度、孔隙率和孔径作为BP神经网络的输入项,用于预测吸声体的平均吸声系数,同时通过调节输入神经元个数、传递函数和隐含层个数构建了最佳的BP神经网络预测模型。对非织造材料基复合吸声体的吸声性能进行预测,并与测试结果进行了对比。结果表明,运用BP神经网络可以建立较理想的适用于复合吸声体平均吸声系数预测的模型。     

植物油的红外光谱结合神经网络快速识别

为实现对植物油的快速检测,借助衰减全反射-傅里叶变换红外光谱分析技术并结合深度学习算法对植物油开展光谱模式识别工作。实验获取8种植物油样本的光谱数据,采用标准正态变换和一阶导数预处理方法消除背景干扰,同时采用竞争性自适应重加权算法模型对各样本特征光谱数据进行提取,分别建立长短记忆神经网络(LSTM)、基于Levenberg-Marquardt算法改进的BP神经网络对提取特征波长后的植物油种类进行预测识别与比较,并采用后者进行了实际样品的识别检测。结果表明,通过提取特征波长,可有效提高LSTM模型的识别准确率,其最优准确率从提取特征波长前的30%～40%提高到80%～90%,模型运行时间从提取特征波长前的111 min 25 s缩短至1 min 45 s。相较于LSTM模型,基于Levenberg-Marquardt算法改进的BP神经网络的分类识别准确率更高,达到99.852%,用于实际样品的识别,识别准确率达到100%。实验结果可为植物油的无损快速检验提供一定的参考与借鉴。     

基于近红外光谱的芝麻油酸价含量的预测

采用近红外光谱分析技术对芝麻油的酸价含量进行检测,避免了传统的化学方法缺陷,同时在不破坏样品的前提下极大地提高了检测效率。对39个芝麻油样本的酸价光谱图进行光谱预处理优化,并选择适当的光谱范围,采用偏最小二乘法(PLS)和BP神经网络算法进行了定量分析研究。结果表明,在所选定的样本和光谱范围内,PLS和BP神经网络算法均可以用于芝麻油酸价含量的预测,采用PLS模型的预测均方根误差(RMSEP)为0.058;用BP神经网络预测的RMSEP为0.148 8,偏最小二乘法建模相对于一般的BP网络建模方法更具有较好的建模预测效果。     

基于人工神经网络的纺粘非织造布孔径及其分布预测

为有效预测纺粘非织造布的孔径及其分布,通过改变计量泵频率和网帘频率制备了30种聚丙烯纺粘非织造布,运用数字图像处理技术测取样品的孔径。以计量泵频率和网帘频率为输入,并通过改变隐含层神经元个数建立了7个BP神经网络模型,对孔径和孔径变异系数进行预测。结果表明,7个模型预测的平均绝对百分比误差均小于5%,其中神经元个数为5的模型的预测精度最高。验证实验的结果进一步印证了BP神经网络模型具有很高的预测准确度。此外BP神经网络模型预测的效果优于多元线性回归模型。     

基于近红外光谱的SSA-RELM的菠萝含水率快速检测

目的：建立快速无损检测菠萝含水率的方法。方法：提出一种基于连续投影法的特征波长选择和麻雀搜索算法（SSA）优化正则化极限学习机（RELM）的菠萝含水率检测模型。针对菠萝近红外光谱数据具有维度高、冗余信息多的特点，分别对比连续投影法、主成分分析法和全波段等筛选特征波长的结果，确定菠萝近红外光谱特征波长筛选方法；针对RELM模型性能受其输入层权值和隐含层偏置的影响，运用麻雀搜索算法优化RELM模型的输入层权值和隐含层偏置，提出一种基于麻雀搜索算法改进正则化极限学习机的菠萝含水率检测模型。结果：与遗传算法改进正则化极限学习机（GA-RELM）、粒子群算法改进正则化极限学习机（PSO-RELM）和RELM相比，基于麻雀算法改进正则化极限学习机（SSA-RELM）的菠萝含水率检测模型的检测精度最高。结论：麻雀搜索算法优化RELM模型可以有效提高RELM模型的菠萝含水率检测精度。     

亚麻织物接枝率与透气率变化关系的数学模型

 摘 要 ： 本设计通过紫外光引发接枝丙烯酸对亚麻织物进行改性处理。应用BP神经网络法和最小二乘回归法分别对不同光照接枝时间、光引发剂用量、单体浓度下的织物接枝率与透气率变化量之间关系进行建模。将接枝率作为输入、透气率变化量作为输出,通过讨论确定神经网络结构为1-10-1,S型函数作为激活函数;同时选出最优网络参数即迭代次数100、训练目标0.001。BP 网络模型与最小二乘模型相比,仿真输出与目标输出相关系数高,误差百分比小。因此,BP神经网络模型具有更好的仿真精度,为接枝率和透气率间关系的探索提供了一种准确有效的预测模型。     

红景天苷缓释微囊神经网络建模

本文在确定输入层和输出层单元、预处理网络数据、选择激活函数、选择训练方法的基础上,建立了红景苷缓释微囊制作参数与性能之间的神经网络模型,网络结构为5-12-3。该模型能较为精确的拟合输入的样本数据,其最大相对误差不超过4%,模型准确可信,可以代替真实试验,该模型的建立为工艺参数的优化打下基础。     

基于BP神经网络的运动文胸肩带属性与乳房振幅的函数关系

为确定运动文胸肩带的3种属性在人体跑步时对胸部振幅的影响,选取8名被测人员,在其左胸上标记6个测量点,更换不同的肩带进行人体运动测试,记录这些测量点动态的三维坐标,进而得到乳房运动的振幅;利用BP神经网络模型,通过更换不同的网络模型参数,确定运动文胸肩带的3种属性与乳房振幅之间的权值关系。结果表明,选取BP神经网络的传输函数为tansig函数,隐含层神经元个数为21个,训练函数为traingdm作为网络参数时,网络拟合出的乳房振幅值达到了真实值的99.44%;在该网络参数下,分别求得网络输入层到隐含层和隐含层到输出层的权值和阈值,最终得到肩带的3种属性与胸部振幅的正向推理关系式。     

基于BP神经网络预测红外-喷动干燥带壳鲜花生水分比

为实现带壳鲜花生红外-喷动干燥过程中水分比的预测,本实验探究了不同干燥温度（55、60、65 ℃和70 ℃）、进口风速（16、17、18 m/s和19 m/s）和助流剂质量（1.0、1.5、2.0 kg和2.5 kg）对带壳鲜花生干燥时间和干燥速率的影响,建立了输入层为干燥温度、进口风速、助流剂质量和干燥时间,隐含层节点数为11,输出层为带壳鲜花生水分比,拓扑结构为“4-11-1”的BP神经网络模型。结果表明：干燥温度和进口风速是影响带壳鲜花生水分比的主要因素,增加进口风速和提高干燥温度能有效缩短带壳鲜花生的干燥时间,提高干燥效率。采用Levenberg-Marquardt（L-M）算法为训练函数,选择tansig-purelin为网络传递函数,经过有限次训练得到的BP神经网络模型,其水分比预测值与实验值之间的决定系数R2为0.99,均方误差为0.02,水分比预测结果相较于传统经典数学模型准确且迅速。本研究建立的BP神经网络模型可为带壳鲜花生在红外-喷动干燥过程中的水分比在线预测提供理论依据和技术支持。     

紫外光谱结合化学计量学检测初榨橄榄油掺伪研究

以紫外光谱为技术手段,结合偏最小二乘法和BP人工神经网络2种化学计量学方法建立了初榨橄榄油/混合橄榄油二元掺伪体系的定量预测模型.试验结果表明,2种统计模型定量预测性能良好,偏最小二乘模型的训练集交叉验证均方根误差RMSEcv和预测集均方根误差RMSEP均达到0.011,预测值与真实值相关性达到0.996 2;BP人工神经网络迭代次数为61步,训练集拟合残差为9.684×10-5,网络预测值和真实值相关系数为0.998 3,对于5％以上掺伪比例的油样BP神经网络能够精确地预测.     

Detection of Escherichia coli in packaged alfalfa sprouts with an electronic nose and an artificial neural network

A rapid method for the detection of Escherichia coli (ATCC 25922) in packaged alfalfa sprouts was developed. Volatile compounds from the headspace of packaged alfalfa sprouts, inoculated with E. coli and incubated at 10 degrees C for 1, 2, and 3 days, were collected and analyzed. Uninoculated sprouts were used as control samples. An electronic nose with 12 metal oxide electronic sensors was used to monitor changes in the composition of the gas phase of the package headspace with respect to volatile metabolites produced by E. coli. The electronic nose was able to differentiate between samples with and without E. coli. To predict the number of E. coli in packaged alfalfa sprouts, an artificial neural network was used, which included an input layer, a hidden layer, and an output layer, with a hyperbolic tangent sigmoidal transfer function in the hidden layer and a linear transfer function in the output layer. The network was shown to be capable of correlating voltametric responses with the number of E. coli. A good prediction was possible, as measured by a regression coefficient (R2 = 0.903) between the actual and predicted data. In conjunction with the artificial neural network, the electronic nose proved to have the ability to detect E. coli in packaged alfalfa sprouts.     

Predicting the drying kinetics of salted codfish (Gadus Morhua): semi‐empirical,diffusive and neural network models

This work aims to compare the accuracy of several drying modelling techniques namely semi‐empirical, diffusive and artificial neural network (ANN) models as applied to salted codfish (Gadus Morhua). To this end, sets of experimental data were collected to adjust parameters for the models. Modelling of codfish drying was performed by resorting to Page and Thompson semi‐empirical models and to a Fick diffusion law. The ANN employed a neural network multilayer ‘feed‐forward’, consisting of one input layer, with four neurons, one hidden layer, formed by five neurons and one output layer with a convergence criterion for training purposes. The simulations showed good results for the ANN (correlation coefficient between 0.987 and 0.999) and semi‐empirical models (correlation coefficient ranging from 0.992 to 0.997 for Page’s model, and from 0.993 to 0.996 for Thompson’s model), while improvements were required to obtain better predictions by the diffusion model (correlation coefficients ranged from 0.864 to 0.959).     

Modelling of solid-phase tea waste extraction for the removal of manganese from food samples by using artificial neural network approach

In this study, a three-layer artificial neural network (ANN) model was employed to develop prediction model for removal of manganese from food samples using tea waste as a low cost adsorbent. After removal of manganese from food samples with acetic acid (5 mol L⁻¹), manganese was adsorbed to a small amount of tea waste, desorbed with nitric acid as a eluent solvent, and determined by flame atomic absorption spectrometry. The input parameters chosen of the model was pH, amount of tea waste, extraction time and eluent concentration. After backpropagation (BP) training, the ANN model was able to predict extraction efficiency of manganese with a tangent sigmoid transfer function at hidden layer and a linear transfer function at output layer. Under the optimum conditions, the detection limit was 0.6 ng g⁻¹. The method was applied to the separation, pre-concentration and determination of manganese in food samples and one reference material.     

Rapid method for prediction of Escherichia coli numbers using an electronic sensor array and an artificial neural network

An electronic sensor array with 12 nonspecific metal oxide sensors was evaluated for its ability to monitor volatile compounds in super broth alone and in super broth inoculated with Escherichia coli (ATCC 25922) at 37 degrees C for 2 to 12 h. Using discriminant function analysis, it was possible to differentiate super broth alone from that containing E. coli when cell numbers were 10(5) CFU or more. There was a good agreement between the volatile profiles from the electronic sensor array and a gas chromatography-mass spectrometer method. The potential to predict the number of E. coli and the concentration of specific metabolic compounds was investigated using an artificial neural network (ANN). The artificial neural network was composed of an input layer, one hidden layer, and an output layer, with a hyperbolic tangent sigmoidal transfer function in the hidden layer and a linear transfer function in the output layer. Good prediction was found as measured by a regression coefficient (R2 = 0.999) between actual and predicted data.     

Solid-phase microextraction, gas chromatography, and mass spectrometry coupled with discriminant factor analysis and multilayer perceptron neural network for detection of Escherichia coli

This study was performed to investigate the ability of using discriminant factor analysis (DFA) and an artificial neural network (ANN) to identify and quantify the number of Escherichia coli (ATCC 25922) in nutrient media from data generated by analysis of E. coli volatile metabolic compounds using solid-phase microextraction (SPME) coupled with gas chromatography (GC) and mass spectrometry (MS). E. coli was grown in super broth and incubated at 37 degrees C for 2 to 12 h. Numbers of E. coli were followed using a colony counting method. An SPME device was used to collect the volatiles from the headspace above the samples, and the volatiles were identified using GC-MS. DFA was used to classify the samples from different incubation times. From DFA, it was possible to differentiate super broth from media containing E. coli when cell numbers were 10(5) CFU or more. The potential to predict the number of E. coli from the SPME-GC-MS data was investigated using a multilayer perceptron (MLP) neural network with back propagation training. The MLP comprised an input layer, one hidden layer, and an output layer, with a hyperbolic tangent sigmoidal transfer function in the hidden layer and a linear transfer function in the output layer. Good prediction was found as measured by a regression coefficient (R2 = 0.996) between actual and predicted data.