A novel nonlinear combination process monitoring method was proposed based on techniques with memory effect (multivariate exponentially weighted moving average (MEWMA)) and kernel independent component analysis (KICA)...A novel nonlinear combination process monitoring method was proposed based on techniques with memory effect (multivariate exponentially weighted moving average (MEWMA)) and kernel independent component analysis (KICA). The method was developed for dealing with nonlinear issues and detecting small or moderate drifts in one or more process variables with autocorrelation. MEWMA charts use additional information from the past history of the process for keeping the memory effect of the process behavior trend. KICA is a recently developed statistical technique for revealing hidden, nonlinear statistically independent factors that underlie sets of measurements and it is a two-phase algorithm: whitened kernel principal component analysis (KPCA) plus independent component analysis (ICA). The application to the fluid catalytic cracking unit (FCCU) simulated process indicates that the proposed combined method based on MEWMA and KICA can effectively capture the nonlinear relationship and detect small drifts in process variables. Its performance significantly outperforms monitoring method based on ICA, MEWMA-ICA and KICA, especially for long-term performance deterioration.展开更多
目的:为实现从母体腹壁混合信号中提取高信噪比和波形清晰的胎儿心电信号,提出一种融合核主成分分析(kernel principal component analysis,KPCA)、快速独立成分分析(fast independent component analysis,FastICA)及奇异值分解(singula...目的:为实现从母体腹壁混合信号中提取高信噪比和波形清晰的胎儿心电信号,提出一种融合核主成分分析(kernel principal component analysis,KPCA)、快速独立成分分析(fast independent component analysis,FastICA)及奇异值分解(singular value decomposition,SVD)的胎儿心电信号提取算法。方法:首先,采用KPCA对母体心电信号进行降维,再利用改进的基于负熵的FastICA处理降维后的数据,得到独立成分。随后,引入样本熵进行信号通道选择,挑选出包含最多母体信息的信号通道。在选中的母体通道上进行SVD,得到母体心电信号的近似估计,再用腹壁源信号减去该信号得到胎儿心电的初步估计。最后,采用改进的基于负熵的FastICA成功分离出纯净的胎儿心电信号。在腹部和直接胎儿心电图数据库(Abdominal and Direct Fetal Electrocardiogram Database,ADFECGDB)和PhysioNet 2013挑战赛数据库中对提出的算法进行验证。结果:提出的算法在主观视觉效果和客观评价指标上都表现出优越的性能。在ADFECGDB数据库中,胎儿QRS复合波检测的敏感度、阳性预测值和F1值分别为99.74%、98.85%和99.30%;在PhysioNet 2013挑战赛数据库中,胎儿QRS复合波检测的敏感度、阳性预测值和F1值分别为99.10%、97.87%和98.48%。结论:融合KPCA、FastICA及SVD的胎儿心电信号提取算法在提取胎儿心电信号的同时有效处理了附加噪声,为胎儿疾病的早期诊断提供了有力支持。展开更多
Investigation of genetic diversity of geographically distant wheat genotypes is </span><span style="font-family:Verdana;">a </span><span style="font-family:Verdana;">useful ...Investigation of genetic diversity of geographically distant wheat genotypes is </span><span style="font-family:Verdana;">a </span><span style="font-family:Verdana;">useful approach in wheat breeding providing efficient crop varieties. This article presents multivariate cluster and principal component analyses (PCA) of some yield traits of wheat, such as thousand-kernel weight (TKW), grain number, grain yield and plant height. Based on the results, an evaluation of economically valuable attributes by eigenvalues made it possible to determine the components that significantly contribute to the yield of common wheat genotypes. Twenty-five genotypes were grouped into four clusters on the basis of average linkage. The PCA showed four principal components (PC) with eigenvalues ></span><span style="font-family:""> </span><span style="font-family:Verdana;">1, explaining approximately 90.8% of the total variability. According to PC analysis, the variance in the eigenvalues was </span><span style="font-family:Verdana;">the </span><span style="font-family:Verdana;">greatest (4.33) for PC-1, PC-2 (1.86) and PC-3 (1.01). The cluster analysis revealed the classification of 25 accessions into four diverse groups. Averages, standard deviations and variances for clusters based on morpho-physiological traits showed that the maximum average values for grain yield (742.2), biomass (1756.7), grains square meter (18</span><span style="font-family:Verdana;">,</span><span style="font-family:Verdana;">373.7), and grains per spike (45.3) were higher in cluster C compared to other clusters. Cluster D exhibited the maximum thousand-kernel weight (TKW) (46.6).展开更多
为了解决多源挥发性有机物(Volatile Organic Compounds,VOCs)数据存在数据维度高、数据关系复杂、数据存在异常的问题,建立了基于核主成分分析(Kernel Principal Component Analysis,KPCA)、孤立森林(Isolated Forest,IF)、加权随机森...为了解决多源挥发性有机物(Volatile Organic Compounds,VOCs)数据存在数据维度高、数据关系复杂、数据存在异常的问题,建立了基于核主成分分析(Kernel Principal Component Analysis,KPCA)、孤立森林(Isolated Forest,IF)、加权随机森林(Weighted Random Forest,WRF)混合方法的VOCs数据清洗模型。首先对研究区域进行网格划分,建立了基于KPCA-IF的VOCs降维异常数据识别模型,通过KPCA方法对多源混合VOCs数据降维,使用IF算法识别异常数据并进行剔除。然后设计了基于WRF的VOCs数据补偿算法,对降维与异常识别后的数据集进行缺失值回归填补。最后,以西安市为例,选取空气质量数据、气象数据等多源VOCs数据进行数据清洗。结果表明,该混合模型可有效对多源VOCs数据降维,进行数据清洗的平均绝对误差为5.08、均方根误差为10.24、中值绝对误差为3.54,均优于对比模型,证明了KPCA-IF-WRF混合模型的鲁棒性更强、精确度更高,具有科学性和可行性。展开更多
基金The National Natural Science Foundation ofChina(No60504033)
文摘A novel nonlinear combination process monitoring method was proposed based on techniques with memory effect (multivariate exponentially weighted moving average (MEWMA)) and kernel independent component analysis (KICA). The method was developed for dealing with nonlinear issues and detecting small or moderate drifts in one or more process variables with autocorrelation. MEWMA charts use additional information from the past history of the process for keeping the memory effect of the process behavior trend. KICA is a recently developed statistical technique for revealing hidden, nonlinear statistically independent factors that underlie sets of measurements and it is a two-phase algorithm: whitened kernel principal component analysis (KPCA) plus independent component analysis (ICA). The application to the fluid catalytic cracking unit (FCCU) simulated process indicates that the proposed combined method based on MEWMA and KICA can effectively capture the nonlinear relationship and detect small drifts in process variables. Its performance significantly outperforms monitoring method based on ICA, MEWMA-ICA and KICA, especially for long-term performance deterioration.
文摘目的:为实现从母体腹壁混合信号中提取高信噪比和波形清晰的胎儿心电信号,提出一种融合核主成分分析(kernel principal component analysis,KPCA)、快速独立成分分析(fast independent component analysis,FastICA)及奇异值分解(singular value decomposition,SVD)的胎儿心电信号提取算法。方法:首先,采用KPCA对母体心电信号进行降维,再利用改进的基于负熵的FastICA处理降维后的数据,得到独立成分。随后,引入样本熵进行信号通道选择,挑选出包含最多母体信息的信号通道。在选中的母体通道上进行SVD,得到母体心电信号的近似估计,再用腹壁源信号减去该信号得到胎儿心电的初步估计。最后,采用改进的基于负熵的FastICA成功分离出纯净的胎儿心电信号。在腹部和直接胎儿心电图数据库(Abdominal and Direct Fetal Electrocardiogram Database,ADFECGDB)和PhysioNet 2013挑战赛数据库中对提出的算法进行验证。结果:提出的算法在主观视觉效果和客观评价指标上都表现出优越的性能。在ADFECGDB数据库中,胎儿QRS复合波检测的敏感度、阳性预测值和F1值分别为99.74%、98.85%和99.30%;在PhysioNet 2013挑战赛数据库中,胎儿QRS复合波检测的敏感度、阳性预测值和F1值分别为99.10%、97.87%和98.48%。结论:融合KPCA、FastICA及SVD的胎儿心电信号提取算法在提取胎儿心电信号的同时有效处理了附加噪声,为胎儿疾病的早期诊断提供了有力支持。
文摘Investigation of genetic diversity of geographically distant wheat genotypes is </span><span style="font-family:Verdana;">a </span><span style="font-family:Verdana;">useful approach in wheat breeding providing efficient crop varieties. This article presents multivariate cluster and principal component analyses (PCA) of some yield traits of wheat, such as thousand-kernel weight (TKW), grain number, grain yield and plant height. Based on the results, an evaluation of economically valuable attributes by eigenvalues made it possible to determine the components that significantly contribute to the yield of common wheat genotypes. Twenty-five genotypes were grouped into four clusters on the basis of average linkage. The PCA showed four principal components (PC) with eigenvalues ></span><span style="font-family:""> </span><span style="font-family:Verdana;">1, explaining approximately 90.8% of the total variability. According to PC analysis, the variance in the eigenvalues was </span><span style="font-family:Verdana;">the </span><span style="font-family:Verdana;">greatest (4.33) for PC-1, PC-2 (1.86) and PC-3 (1.01). The cluster analysis revealed the classification of 25 accessions into four diverse groups. Averages, standard deviations and variances for clusters based on morpho-physiological traits showed that the maximum average values for grain yield (742.2), biomass (1756.7), grains square meter (18</span><span style="font-family:Verdana;">,</span><span style="font-family:Verdana;">373.7), and grains per spike (45.3) were higher in cluster C compared to other clusters. Cluster D exhibited the maximum thousand-kernel weight (TKW) (46.6).
文摘为了解决多源挥发性有机物(Volatile Organic Compounds,VOCs)数据存在数据维度高、数据关系复杂、数据存在异常的问题,建立了基于核主成分分析(Kernel Principal Component Analysis,KPCA)、孤立森林(Isolated Forest,IF)、加权随机森林(Weighted Random Forest,WRF)混合方法的VOCs数据清洗模型。首先对研究区域进行网格划分,建立了基于KPCA-IF的VOCs降维异常数据识别模型,通过KPCA方法对多源混合VOCs数据降维,使用IF算法识别异常数据并进行剔除。然后设计了基于WRF的VOCs数据补偿算法,对降维与异常识别后的数据集进行缺失值回归填补。最后,以西安市为例,选取空气质量数据、气象数据等多源VOCs数据进行数据清洗。结果表明,该混合模型可有效对多源VOCs数据降维,进行数据清洗的平均绝对误差为5.08、均方根误差为10.24、中值绝对误差为3.54,均优于对比模型,证明了KPCA-IF-WRF混合模型的鲁棒性更强、精确度更高,具有科学性和可行性。
