基于BP神经网络的猪肉新鲜度检测方法

猪肉在储藏、加工和运输过程中因为腐败会挥发出氨气、硫化氢等,据此选择8个金属氧化物半导体气敏传感器构成检测阵列,运用改进的BP神经网络算法建立猪肉新鲜度智能检测的数学模型,从而构建了猪肉新鲜度检测电子鼻系统。通过检测实验构建样本数据集,并对识别模型进行训练、测试,结果表明该模型对猪肉新鲜度的预测结果与用理化分析方法所得实际结果具有很好的吻合度,预测准确率大于90%。     

电子鼻判别小麦陈化年限的检测方法研究

采用电子鼻对五个储藏年限的陈化小麦进行年限分析,确定了采用电子鼻判别小麦储藏年限的最佳参数及方法.对传感器信号进行多因素方差分析可知:对于固定容器的陈化小麦样品,不同的小麦密封时间对电子鼻的响应信号的影响极为显著;其次是小麦在烧杯内的密封质量.通过静置密封时间和密封质量的方差分析,得出小麦在500 mL烧杯内的最佳静置时间为1.5 h,密封在烧杯内的小麦最恰当质量为50 g.采用以上参数,对五个储藏年限的小麦进行辨别,PCA分析可以将不同储藏年限的小麦较好的区分开来,并且五个年份的小麦自右上角至左下角依次排列;而LDA分析能够将差别年限较大的陈化小麦进行区分,差距较小的,不能够很好的区分,其区分效果不如PCA分析;进而采用BP神经网络的方法进行判别分析,训练样本正确率为100%,测试样本正确率也达到了85%以上.     

电子鼻技术对甘薯储藏中长喙壳菌侵染病变程度判别

针对甘薯受到病原菌侵害后，引发储藏病害产生毒素问题，提出了一种基于电子鼻技术，利用化学计量学对甘薯储藏病害的病变程度进行判别的方法，以期为库存甘薯储藏病害的识别提供技术参考。通过主成分分析(PCA)、偏最小二乘法 －判别分析(PLS-DA)、正交偏最小二乘法判别分析(OPLS-DA)、BP神经网络(BPNN)和支持向量机(SVM)对长喙壳菌侵染后不同病变程度的甘薯气味响应特征值进行分析判别。结果表明，PCA、PLS-DA、OPLS-DA)均可对3类病变程度的甘薯进行有效的区分。BPNN 模型未能对3类病变甘薯进行有效的区分。SVM利用交叉验证算法优化(Best c＝0.2500 Best g＝0.3789)并建立模型。测试集的准确率为96%。因此，电子鼻技术可对甘薯储藏病害程度实现区分，且具有较好的判别效果。     

基于电子鼻的香蕉储存时间鉴别方法研究

研究了一种基于电子鼻系统的香蕉储存时间鉴别方法。实验检测了不同储存时间的香蕉样品,主成分分析方法可以较好地区分不同储存时间的香蕉样品,同时检验了样品的微生物指标以探讨电子鼻响应与微生物指标之间的关系。随机共振信噪比谱不但可以区分香蕉样品,同时基于信噪比特征值建立的香蕉储存时间鉴别模型具有较高的预测准确率。该方法具有较好的实际应用价值。     

基于电子鼻的芒果储存时间预测方法研究

本文研究了一种芒果储存期预测方法,使用智能电子鼻实验检测了存储于9天内的芒果样品,主成分分析法实现了不同贮存时间芒果样品的区分,采用阈值随机共振方法提取芒果品质特性信息,并以互相关系数极大值构建芒果储存期预测模型。预测实验结果表明该模型预测准确度为87.5%。该预测方法具有检测快速、准确性好、成本低等优势。     

基于金属氧化物传感器阵列的小麦霉变程度检测

研制了一套由8个金属氧化物传感器组成、用于检测小麦霉变的电子鼻系统.使用该电子鼻对不同霉变程度和掺入不同百分比含量霉麦的小麦样品进行检测.通过方差分析和主成分分析优化传感器阵列并去掉冗余传感器,对优化后的数据进行主成分分析(PCA)和线性判别分析(LDA),其中PCA的前两个主成分对两类实验结果分析的总贡献率为98.30％和99.27％,LDA前两个判别因子对两类实验结果分析的总贡献率为99.68％和93.30％,且由得分图可知两种方法均能很好地区分不同的小麦样品.利用BP神经网络建立预测模型,对样品菌落总数和掺入样品中霉麦的百分比进行预测.两种预测模型的预测值和测量值之间的相关系数分别为0.91和0.94,表明预测模型具有较好预测性能.     

电子鼻信号特征提取与传感器优化的研究

采用PEN2型电子鼻系统对芝麻油的玉米油掺假进行定性鉴别和定量预测,运用主成分分析,逐步判别分析和Fisher线性判别函数变换对原始数据进行预处理,从而降低原始数据空间的维数,并用判别分析与人工神经网络对数据进行进一步分析,考察了不同的数据预处理方法的效果.判别分析结果表明,采用Fisher线性判别函数变换所得到的十个变量判别能力最强,误判率为0.61%,仅有1个样品出现误判.在BP神经网络的定量预测中,采用逐步判别分析所筛选出的十个变量作为网络输入,所得的预测结果最为理想,绝对误差个体值的95%置信区间最小,为(-4.71%,3.38%),均方误差为4.75,预测值与实际值之间有极显著的相关性,相关系数R=0.998 08.     

基于气敏传感器阵列和PCA的猪肉新鲜度快速分类方法

根据猪肉的腐败机理及其气味特征,合理选用了气敏传感器阵列,建立了一套用于猪肉新鲜度识别的电子鼻系统。通过不同保存温度不同时间的猪肉样品的电子鼻检测实验,探讨了电子鼻的实时响应特性的补偿方法和基于PCA的猪肉新鲜度的判别模式。同时采用挥发性盐基氮(TVBN)和微生物菌落总数检测实验进行对比分析。结果表明:在不同实验条件下,由于微生物作用产生的猪肉腐败规律,可由电子鼻实时检测;把温湿度信息作为PCA输入分量,能较好补偿温湿度造成的传感器误差。但是对于不同保存温度,判别模式是不同的。结果也表明,建立的电子鼻系统可以分析腐败过程,并半定量地表征猪肉腐败水平。     

基于Seq2Seq和Attention的时序卫星云图台风等级预测

台风预测可为台风预警预报提供先验信息,辅助相关部门进行科学决策,以减少灾害损失。利用时间序列台风卫星云图,提出一种新的台风等级预测模型SeqTyphoon,将注意力机制和序列到序列引入模型预测未来时刻台风图像,然后利用卷积神经网络对预测的台风图像进行台风等级预测。通过日本气象厅发布的1981—2017年3万多张时序台风卫星云图,构建了训练集、验证集和测试集,分别对应29 519、3 804、1 995张台风图像。针对SeqTyphoon模型,分别进行了台风云图的不同时间间隔、不同预测时长及不同空间分辨率对台风图像预测精度影响的对比实验。实验结果表明,台风云图均为32像素×32像素,时间间隔为6h比时间间隔为12h的训练集和验证集的均方根误差分别降低5.41%、5.72%,前者训练集的均方根误差达到0.092 2,验证集为0.095 4,前者台风等级预测准确率为后者的2倍;台风云图为32像素×32像素,时间间隔为6h时,预测未来6～48h的台风图像,训练集和验证集的均方根误差均递增,台风等级预测准确率递减;时间间隔为6h,图像为64像素×64像素的训练集的均方根误差为0.089 6,验证集为0.091 1,台风等级预测总体准确率为83.2%。综上,影响台风图像的最主要因素是相邻台风云图的时间间隔,其次是预测时长与空间分辨率大小。     

使用电子鼻检测低温贮藏罗非鱼储存时间

本文采用电子鼻结合理化检验方法探索了一种低温贮藏罗非鱼储存时间预测方法。按照国家标准检验了罗非鱼样品的挥发性盐基氮（TVBN）,同时测量了电子鼻传感器阵列响应,以TVBN检验结果标定罗非鱼的新鲜度。在随机共振理论模型研究的基础之上,对电子鼻检测数据进行主成分和随机共振分析。相对于主成分分析结果,随机共振输出信噪比可以完全区分罗非鱼样品。依据TVBN国家标准计算得到罗非鱼电子鼻检测信噪比新鲜度阈值为-61.1688。选取信噪比曲线特征值经线性拟合回归建立了罗非鱼储存时间预测模型,该模型的预测系数R2=0.910,结果表明可以准确预测罗非鱼的储存时间。该方法具有快速、易操作、准确等优势,有望于在水产品品质快速分析中得到广泛应用。     

基于电子鼻的番茄种子不同储藏时间的鉴别研究

采用电子鼻对三种不同年份的番茄种子进行分析.结果表明:利用电子鼻可以很好的区分不同年份的番茄种子;利用主成分分析方法(PCA)基本上可以辨别出不同掺杂比例的种子,但是当掺杂比例为37.5%和50%时,较难利用电子鼻进行辨别区分;利用线性判别分析方法(LDA)可以很好的辨别出不同掺杂比例的番茄种子,并且每个混合种类的区域...     

Multi-view Incremental Discriminant Analysis

In recent years, the use of multi-view data has attracted much attention resulting in many multi-view batch learning algorithms. However, these algorithms prove expensive in terms of training time and memory when used on the incremental data. In this paper, we propose Multi-view Incremental Discriminant Analysis (MvIDA), which updates the trained model to incorporate new data samples. MvIDA requires only the old model and newly added data to update the model. Depending on the nature of the increments, MvIDA is presented as two cases, sequential MvIDA and chunk MvIDA. We have compared the proposed method against the batch Multi-view Discriminant Analysis (MvDA) for its discriminability, order independence, the effect of the number of views, training time, and memory requirements. We have also compared our method with single-view Incremental Linear Discriminant Analysis (ILDA) for accuracy and training time. The experiments are conducted on four datasets with a wide range of dimensions per view. The results show that through order independence and faster construction of the optimal discriminant subspace, MvIDA addresses the issues faced by the batch multi-view algorithms in the incremental setting.     

An electronic nose system based on a micro-machined gas sensor array to assess the freshness of sardines

An electronic nose system based on a four-element, integrated, micro-machined, metal oxide gas sensor array is used to assess, in an objective manner, the evolutionary stages of freshness in sardine samples stored up to 1-week at 4 °C. The sensors developed were based on tin oxide doped with Pt or Pd or Bi, and on tungsten oxide doped with Au. The selection of the gas sensitive materials was based on a previous identification and quantification of characteristic compounds found in the headspace of sardines determined by solid phase micro-extraction gas chromatography coupled to mass spectrometry. Principal component analysis performed on the responses of the sensor array revealed that sardine samples could be classified in three freshness states. This was in good agreement with the results of a microbiological analysis. A support vector machine-based classifier reached a 100% success rate in the identification of sardine freshness. The stability of the electronic nose classification ability was assessed by correctly classifying measurement databases gathered 1-month apart. By building and validating quantitative partial least squares models, which employed as input data the gas sensor responses, it was possible to predict with good accuracy the total viable counts (TVC) of aerobic bacteria present in sardine samples. For the validation dataset, the correlation coefficient between actual and predicted TVC was 0.91, which indicates that the electronic nose system developed is a simple and rapid technique for evaluating sardine freshness.     

Aging fingerprint characterization of beer using electronic nose

In this work, attempts were made in order to characterize the change of aroma of alcoholic and non alcoholic beers during the aging process by use of a metal oxide semiconductor based electronic nose. The aged beer samples were statistically characterized in several classes. Linear techniques as principal component analysis (PCA) and Linear Discriminant Analaysis (LDA) were performed over the data that revealed non alcoholic beer classes are separated except a partial overlapping between zones corresponding to two specified classes of the aged beers. A clear discrimination was not found among the alcoholic beer classes showing the more stability of such type of beer compared with non alcoholic beer. In this research, to classify the classes, two types of artificial neural networks were used: Probabilistic Neural Networks (PNN) with Radial Basis Functions (RBF) and FeedForward Networks with Backpropagation (BP) learning method. The classification success was found to be 90% and 100% for alcoholic and non alcoholic beers, respectively. Application of PNN showed the classification accuracy of 83% and 100%, respectively for the aged alcoholic and non alcoholic beer classes as well. Finally, this study showed the capability of the electronic nose system for the evaluation of the aroma fingerprint changes in beer during the aging process.     

Classification of non-alcoholic beer based on aftertaste sensory evaluation by chemometric tools

Sensory evaluation is the application of knowledge and skills derived from several different scientific and technical disciplines, physiology, chemistry, mathematics and statistics, human behavior, and knowledge about product preparation practices. This research was aimed to evaluate aftertaste sensory attributes of commercial non-alcoholic beer brands (P1, P2, P3, P4, P5, P6, P7) by several chemometric tools. These attributes were bitter, sour, sweet, fruity, liquorice, artificial, body, intensity and duration. The results showed that the data are in a good consistency. Therefore, the brands were statistically classified in several categories. Linear techniques as Principal Component Analysis (PCA) and Linear Discriminant Analysis (LDA) were performed over the data that revealed all types of beer are well separated except a partial overlapping between zones corresponding to P4, P6 and P7. In this research, for the confirmation of the groups observed in PCA and in order to calculate the errors in calibration and in validation, PLS-DA technique was used. Based on the quantitative data of PLS-DA, the classification accuracy values were ranked within 49-86%. Moreover, it was found that the classification accuracy of LDA was much better than PCA. It shows that this trained sensory panel can discriminate among the samples except an overlapping between two types of beer. Also, two types of artificial networks were used: Probabilistic Neural Networks (PNN) with Radial Basis Functions (RBF) and FeedForward Networks with Back Propagation (BP) learning method. The highest classification success rate (correct predicted number over total number of measurements) of about 97% was obtained for RBF followed by 94% for BP. The results obtained in this study could be used as a reference for electronic nose and electronic tongue in beer quality control.