A novel nonlinear process monitoring and fault detection method based on kernel independent component analysis(ICA) is proposed.The kernel ICA method is a two-phase algorithm:whitened kernel principal component(KPCA) ...A novel nonlinear process monitoring and fault detection method based on kernel independent component analysis(ICA) is proposed.The kernel ICA method is a two-phase algorithm:whitened kernel principal component(KPCA) plus ICA.KPCA spheres data and makes the data structure become as linearly separable as possible by virtue of an implicit nonlinear mapping determined by kernel.ICA seeks the projection directions in the KPCA whitened space,making the distribution of the projected data as non-gaussian as possible.The application to the fluid catalytic cracking unit(FCCU) simulated process indicates that the proposed process monitoring method based on kernel ICA can effectively capture the nonlinear relationship in process variables.Its performance significantly outperforms monitoring method based on ICA or KPCA.展开更多
对农作物品种正确分类是作物分类学的重要内容,为考察X-ray成像技术对小麦品种分类研究的有效性,基于软X-ray成像仪采集的3品种(Kama,Rosa and Canadian)每个品种70个籽粒,共210个籽粒样本的X-ray扫描图像,并针对其7个形态几何特征(面...对农作物品种正确分类是作物分类学的重要内容,为考察X-ray成像技术对小麦品种分类研究的有效性,基于软X-ray成像仪采集的3品种(Kama,Rosa and Canadian)每个品种70个籽粒,共210个籽粒样本的X-ray扫描图像,并针对其7个形态几何特征(面积、周长、紧致度、籽粒长度、宽度、偏斜度、种子腹沟长度),提出了一种使用Kernel-ICA的方法先对特征进行优化,再进行小麦品种的聚类与识别的方法,并与K-means、C-means 2种聚类方法以及基于工神经网络(ANN)和支持向量机(SVM)2种识别方法的分类结果进行比较,结果发现:分类正确率从高到低分别为:Kernel-ICA、SVM、C-means、K-means、BP-ANN,分类正确率分别为:91.9%、90.5%、89.5%、87.1%、86.9%。研究提出的Kernel-ICA的方法,聚类优化和识别能力较强,对软X-ray成像的小麦品种进行分类,已基本上满足农艺上对小麦品种分类需要,对农作物种质资源鉴别和作物品种分类研究具有积极意义。展开更多
核主成分分析(Kernel principal component analysis,KPCA)是一种非线性降维工具,在降低数据流分类处理量方面发挥着积极作用.然而,由于复杂性太高,导致KPCA的降维能力有限.为此,本文给出了一种增量核主成分分析算法(Incremental KPCA f...核主成分分析(Kernel principal component analysis,KPCA)是一种非线性降维工具,在降低数据流分类处理量方面发挥着积极作用.然而,由于复杂性太高,导致KPCA的降维能力有限.为此,本文给出了一种增量核主成分分析算法(Incremental KPCA for dimensionality-reduction,IKDR),该算法在每步迭代估计中只需线性内存开销,大大降低了复杂性.在IKDR的基础上,结合BP(Back propagation)神经网络提出了数据流在线分类框架:IKOCFrame(Online classificationframe based on IKDR).通过一系列真实和人工数据集上的实验,检验了IKDR算法的收敛性,并且验证了IKOCFrame相对于同类基于成分分析的分类算法的优越性.展开更多
基金National Nature Science Foundation of China (No60504033)
文摘A novel nonlinear process monitoring and fault detection method based on kernel independent component analysis(ICA) is proposed.The kernel ICA method is a two-phase algorithm:whitened kernel principal component(KPCA) plus ICA.KPCA spheres data and makes the data structure become as linearly separable as possible by virtue of an implicit nonlinear mapping determined by kernel.ICA seeks the projection directions in the KPCA whitened space,making the distribution of the projected data as non-gaussian as possible.The application to the fluid catalytic cracking unit(FCCU) simulated process indicates that the proposed process monitoring method based on kernel ICA can effectively capture the nonlinear relationship in process variables.Its performance significantly outperforms monitoring method based on ICA or KPCA.
文摘对农作物品种正确分类是作物分类学的重要内容,为考察X-ray成像技术对小麦品种分类研究的有效性,基于软X-ray成像仪采集的3品种(Kama,Rosa and Canadian)每个品种70个籽粒,共210个籽粒样本的X-ray扫描图像,并针对其7个形态几何特征(面积、周长、紧致度、籽粒长度、宽度、偏斜度、种子腹沟长度),提出了一种使用Kernel-ICA的方法先对特征进行优化,再进行小麦品种的聚类与识别的方法,并与K-means、C-means 2种聚类方法以及基于工神经网络(ANN)和支持向量机(SVM)2种识别方法的分类结果进行比较,结果发现:分类正确率从高到低分别为:Kernel-ICA、SVM、C-means、K-means、BP-ANN,分类正确率分别为:91.9%、90.5%、89.5%、87.1%、86.9%。研究提出的Kernel-ICA的方法,聚类优化和识别能力较强,对软X-ray成像的小麦品种进行分类,已基本上满足农艺上对小麦品种分类需要,对农作物种质资源鉴别和作物品种分类研究具有积极意义。
文摘核主成分分析(Kernel principal component analysis,KPCA)是一种非线性降维工具,在降低数据流分类处理量方面发挥着积极作用.然而,由于复杂性太高,导致KPCA的降维能力有限.为此,本文给出了一种增量核主成分分析算法(Incremental KPCA for dimensionality-reduction,IKDR),该算法在每步迭代估计中只需线性内存开销,大大降低了复杂性.在IKDR的基础上,结合BP(Back propagation)神经网络提出了数据流在线分类框架:IKOCFrame(Online classificationframe based on IKDR).通过一系列真实和人工数据集上的实验,检验了IKDR算法的收敛性,并且验证了IKOCFrame相对于同类基于成分分析的分类算法的优越性.
