Monitoring high-dimensional multistage processes becomes crucial to ensure the quality of the final product in modern industry environments. Few statistical process monitoring(SPC) approaches for monitoring and contro...Monitoring high-dimensional multistage processes becomes crucial to ensure the quality of the final product in modern industry environments. Few statistical process monitoring(SPC) approaches for monitoring and controlling quality in highdimensional multistage processes are studied. We propose a deviance residual-based multivariate exponentially weighted moving average(MEWMA) control chart with a variable selection procedure. We demonstrate that it outperforms the existing multivariate SPC charts in terms of out-of-control average run length(ARL) for the detection of process mean shift.展开更多
Polychlorinated dibenzothiophenes（PCDTs） are a group of important persistent organic pollutants.In the present study,geometrical optimization and electrostatic potential calculations have been performed for all 135 ...Polychlorinated dibenzothiophenes（PCDTs） are a group of important persistent organic pollutants.In the present study,geometrical optimization and electrostatic potential calculations have been performed for all 135 PCDTs congeners at the B3LYP/6-31G＊ level of theory.By means of the VSMP（variable selection and modeling based on prediction） program,one optimal descriptor（molecular polarizability,α） was selected to develop a QSRR model for the prediction of gas chromatographic retention indices（GC-RI） of PCDTs.The estimated correlation coefficients（r2） and LOO-validated correlation coefficients（q2）,all more than 0.99,were built by multiple linear regression,which shows a good estimation ability and stability of the models.A prediction power for the external samples was validated by the model built from the training set with 17 polychlorinated dibenzothiophenes.展开更多
Government credibility is an important asset of contemporary national governance, an important criterion for evaluating government legitimacy, and a key factor in measuring the effectiveness of government governance. ...Government credibility is an important asset of contemporary national governance, an important criterion for evaluating government legitimacy, and a key factor in measuring the effectiveness of government governance. In recent years, researchers’ research on government credibility has mostly focused on exploring theories and mechanisms, with little empirical research on this topic. This article intends to apply variable selection models in the field of social statistics to the issue of government credibility, in order to achieve empirical research on government credibility and explore its core influencing factors from a statistical perspective. Specifically, this article intends to use four regression-analysis-based methods and three random-forest-based methods to study the influencing factors of government credibility in various provinces in China, and compare the performance of these seven variable selection methods in different dimensions. The research results show that there are certain differences in simplicity, accuracy, and variable importance ranking among different variable selection methods, which present different importance in the study of government credibility issues. This study provides a methodological reference for variable selection models in the field of social science research, and also offers a multidimensional comparative perspective for analyzing the influencing factors of government credibility.展开更多
Ballistic missile defense system (BMDS) is important for its special role in ensuring national security and maintaining strategic balance. Research on modeling and simulation of the BMDS beforehand is essential as dev...Ballistic missile defense system (BMDS) is important for its special role in ensuring national security and maintaining strategic balance. Research on modeling and simulation of the BMDS beforehand is essential as developing a real one requires lots of manpower and resources. BMDS is a typical complex system for its nonlinear, adaptive and uncertainty characteristics. The agent-based modeling method is well suited for the complex system whose overall behaviors are determined by interactions among individual elements. A multi-agent decision support system (DSS), which includes missile agent, radar agent and command center agent, is established based on the studies of structure and function of BMDS. Considering the constraints brought by radar, intercept missile, offensive missile and commander, the objective function of DSS is established. In order to dynamically generate the optimal interception plan, the variable neighborhood negative selection particle swarm optimization (VNNSPSO) algorithm is proposed to support the decision making of DSS. The proposed algorithm is compared with the standard PSO, constriction factor PSO (CFPSO), inertia weight linear decrease PSO (LDPSO), variable neighborhood PSO (VNPSO) algorithm from the aspects of convergence rate, iteration number, average fitness value and standard deviation. The simulation results verify the efficiency of the proposed algorithm. The multi-agent DSS is developed through the Repast simulation platform and the constructed DSS can generate intercept plans automatically and support three-dimensional dynamic display of missile defense process.展开更多
Using a novel freshwater photobacteria — Q67 as an indication organism and the VeritasTM luminometer with 96-well microplate as the testing equipment to determine luminous intensity of photobacteria,the familiar 29 s...Using a novel freshwater photobacteria — Q67 as an indication organism and the VeritasTM luminometer with 96-well microplate as the testing equipment to determine luminous intensity of photobacteria,the familiar 29 substituted benzenes of the median inhibition toxicities（pEC50）were determined,respectively.The quantum chemical parameters of 29 substituted benzenes in the ideal gas state at 298.15 K and 1.013×105 Pa have been calculated at the B3LYP/6-31G＊ level using Gaussian 03 program.The Quantitative linear relationship（N1）between the pEC50 and two descriptors of 29 substituted benzenes was developed using the variable selection and modeling based on prediction（VSMP）.Model N1 showed good estimation ability and stability（r = 0.8777,q = 0.8482）,which exhibited the difference between empirical and predicted values of 2,3-dimethylphenol was greater（0.5）,so it was given up.Using VSMP to select the optimal descriptors,a 2-variable multiple linear regression model（called model N2）was developed for the pEC50 of substituted benzenes.The r and q for model N2 based on 28 substituted benzenes are 0.8991 and 0.8735,respectively.In order to validate the model,28 substituted benzenes were divided into a training set consisting of 20 compounds and a test set with 8 compounds.The result showed that some main structural factors influencing the pEC50 of substituted benzenes are the lowest unoccupied orbital（ELUMO）and total energy（EHF）.展开更多
变量预测模型的模式识别方法(Variable predictive model based class discriminate,VPMCD)是一种利用特征值相互内在关系进行模式识别的新方法。论文提出了基于局部均值分解LMD(Local mean decomposition,LMD)能量矩概念,并针对轴承故...变量预测模型的模式识别方法(Variable predictive model based class discriminate,VPMCD)是一种利用特征值相互内在关系进行模式识别的新方法。论文提出了基于局部均值分解LMD(Local mean decomposition,LMD)能量矩概念,并针对轴承故障振动信号特征值的相互内在联系,将LMD能量矩与变量预测模型模式识别相结合,提出了一种轴承故障智能诊断新方法。首先利用LMD方法将复杂非平稳的原始信号分解为若干PF(Product function,PF)分量;然后利用相关分析剔除LMD方法中的虚假PF分量,并提取真实PF分量能量矩组成特征向量来有效地表达故障信息;最后采用VPMCD方法进行轴承故障诊断。通过仿真信号验证了PF能量矩比PF能量更能反映非平稳信号本质特征。轴承故障诊断实验结果表明,论文提出的方法能有效地应用于小样本多分类轴承故障智能诊断。展开更多
提出了基于VPMCD(Variable Predictive Model Based Class Discriminate,简称VPMCD)和EMD(Empirical mode decomposition,简称EMD)的齿轮故障诊断方法,并将它应用于齿轮稳态信号的分析。VPMCD方法是一种新的模式识别方法,特别适合于非...提出了基于VPMCD(Variable Predictive Model Based Class Discriminate,简称VPMCD)和EMD(Empirical mode decomposition,简称EMD)的齿轮故障诊断方法,并将它应用于齿轮稳态信号的分析。VPMCD方法是一种新的模式识别方法,特别适合于非线性分类问题,它充分利用从原始数据中所提取的特征值之间的相互内在关系建立数学模型,从而进行模式识别。在基于VPMCD和EMD的齿轮故障诊断方法中,首先采用EMD方法将齿轮振动信号自适应地分解为若干个单分量信号,然后提取各个分量的样本熵并将其作为特征值,最后采用VPMCD分类器进行故障识别和分类。结果表明该方法能够有效地突出齿轮故障振动信号的故障特征,提高了齿轮故障诊断的准确性。展开更多
针对滚动轴承故障诊断时所提取的特征值中可能含有较小相关性和冗余性特征,采用基于Wrapper模式的距离评价技术(distance evaluation technique,简称DET)进行特征选择。在分类器的设计中,提出了基于稳健回归的多变量预测模型(Robust reg...针对滚动轴承故障诊断时所提取的特征值中可能含有较小相关性和冗余性特征,采用基于Wrapper模式的距离评价技术(distance evaluation technique,简称DET)进行特征选择。在分类器的设计中,提出了基于稳健回归的多变量预测模型(Robust regression-Variable predictive model based class discriminate,简称RRVPMCD)分类方法,以减小"异常值"对参数估计的影响,从而有望建立更加准确的预测模型。即根据Wrapper模式的特点,首先通过DET方法计算出各特征值对类的敏感度,并结合RRVPMCD分类器,选择敏感度最大的若干特征值组成特征向量矩阵;然后用RRVPMCD方法进行训练,建立预测模型;最后用所建立的预测模型进行模式识别。实验分析结果表明,基于Wrapper模式的特征选择方法和RRVPMCD分类方法相结合可以有效地对滚动轴承的工作状态和故障类型进行识别。展开更多
提出一种基于拉普拉斯特征映射流形学习算法(Laplacian Eigenmaps,简称LE)和改进多变量预测模型(Variable predictive model based class discriminate,简称VPMCD)的滚动轴承故障诊断方法,首先对振动信号进行局部特征尺度分解(Local cha...提出一种基于拉普拉斯特征映射流形学习算法(Laplacian Eigenmaps,简称LE)和改进多变量预测模型(Variable predictive model based class discriminate,简称VPMCD)的滚动轴承故障诊断方法,首先对振动信号进行局部特征尺度分解(Local characteristic scale decomposition,简称LCD),并提取各内禀尺度分量(Intrinsic scale component,简称ISC)的特征构造高维特征向量,接着采用LE算法挖掘出高维数据中包含有效信息且具有内在规律性的低维特征,然后输入到基于Kriging的改进多变量预测模型(Kriging-variable predictive model based class discriminate,简称KVPMCD)分类器中进行模式识别。该方法充分利用并有效结合了LCD在信号处理、LE在挖掘特征信息和KVPMCD在模式识别方面的优势,实现了滚动轴承故障特征提取到故障识别的全程诊断。实验分析结果表明:基于LE算法和KVPMCD的分类方法可以有效地对滚动轴承的工作状态和故障类型进行识别。展开更多
VPMCD(Variable Predictive Model Based Class Discriminate)是一种新的模式识别方法,它充分利用从原始数据中所提取的特征值之间的相互内在关系建立数学模型,从而进行模式识别。论文将VPMCD结合排列熵(Permutation Entropy,简称PE)方...VPMCD(Variable Predictive Model Based Class Discriminate)是一种新的模式识别方法,它充分利用从原始数据中所提取的特征值之间的相互内在关系建立数学模型,从而进行模式识别。论文将VPMCD结合排列熵(Permutation Entropy,简称PE)方法应用于滚动轴承故障诊断。首先采用ITD(Intrinsic Time-scale Decomposition,简称ITD)对滚动轴承振动信号进行分解,得到若干个固有旋转(Proper Rotation)分量,并对包含主要故障信息的PR分量提取排列熵作为故障特征值;然后,对VPMCD分类器进行训练;最后,采用VPMCD分类器进行故障识别和分类。实验数据的分析结果表明该方法能够有效地应用于滚动轴承故障诊断。展开更多
基于变量预测模型的分类识别(Variable predictive model-based class discriminate,VPMCD)方法是一种新的分类识别方法,但模型类型的选择存在主观性。为了解决VPMCD方法应用于机械故障诊断过程中的模型选择问题,结合遗传算法的全局优...基于变量预测模型的分类识别(Variable predictive model-based class discriminate,VPMCD)方法是一种新的分类识别方法,但模型类型的选择存在主观性。为了解决VPMCD方法应用于机械故障诊断过程中的模型选择问题,结合遗传算法的全局优化能力,提出了基于GA-VPMCD(Genetic algorithm and variable predictive model based class discriminate)智能诊断方法。首先通过样本训练建立多个弱VPM(Variable predictive model),然后采用遗传算法优化各个弱VPM的权值,得到最优权值矩阵,最后用最优权值矩阵加权融合测试样本的弱VPM特征变量预测值,得到最佳特征变量预测值,并以误差平方和最小为辨别函数分类识别故障类型。通过GA-VPMCD方法在滚动轴承故障智能诊断中的应用实验验证了基于GA-VPMCD的故障智能诊断方法能有效地提高诊断精度和诊断系统的鲁棒性。展开更多
基于变量预测模型的模式识别(variable predictive model based class discriminate,VPMCD)方法是一种充分利用特征值之间相互内在关系进行多分类模式识别的新方法。对VPMCD算法进行了研究,并采用交叉验证法来选择VPMCD模型。针对机械...基于变量预测模型的模式识别(variable predictive model based class discriminate,VPMCD)方法是一种充分利用特征值之间相互内在关系进行多分类模式识别的新方法。对VPMCD算法进行了研究,并采用交叉验证法来选择VPMCD模型。针对机械故障振动信号的特征值之间的相互内在关系,结合本征时间尺度分解(intrinsic time-scale decom-position,ITD),提出了一种基于本征时间尺度分解和VPMCD的机械故障诊断方法。该方法首先利用ITD方法将原始信号分解若干个PR(proper rotation,PR)分量,然后提取第一个PR分量的无量纲时域统计参数组成特征向量,最后采用VPMCD方法进行机械故障诊断。通过滚动轴承故障诊断实验验证了该方法能有效地应用于小样本多分类机械故障诊断。展开更多
将基于变量预测模型的模式识别(variable predictive model based class discriminate,简称VPMCD)、独立分量分析(independent component analysis,简称ICA)和相关系数分析方法相结合,提出了基于ICA相关系数和VPMCD的滚动轴承故障诊断...将基于变量预测模型的模式识别(variable predictive model based class discriminate,简称VPMCD)、独立分量分析(independent component analysis,简称ICA)和相关系数分析方法相结合,提出了基于ICA相关系数和VPMCD的滚动轴承故障诊断方法。首先,对不同工况下的滚动轴承振动信号分别进行独立分量分析,获得各工况信号的独立分量;然后,提取样本与不同工况信号独立分量之间的相关系数,并以相关系数绝对值的和作为该样本的特征值;最后,采用VPMCD分类器进行故障识别和分类。实验数据的分析结果表明,该方法能够有效应用于滚动轴承故障诊断。展开更多
基金supported by the Qatar National Research Fund(NPRP5-364-2-142NPRP7-1040-2-293)
文摘Monitoring high-dimensional multistage processes becomes crucial to ensure the quality of the final product in modern industry environments. Few statistical process monitoring(SPC) approaches for monitoring and controlling quality in highdimensional multistage processes are studied. We propose a deviance residual-based multivariate exponentially weighted moving average(MEWMA) control chart with a variable selection procedure. We demonstrate that it outperforms the existing multivariate SPC charts in terms of out-of-control average run length(ARL) for the detection of process mean shift.
基金Sponsored by the NSF of Guangxi Province (No. 2011XNSFA018059)Guangxi Key Laboratory Research Fund of Environmental Engineering and Protection Assessment (No. 0801Z026)+1 种基金Major Science of Water Pollution Control and Management (No. 2008ZX07317-02)the Guangxi Zhuang Autonomous Region Department of Education Research (No. 201010LX174) Funding
文摘Polychlorinated dibenzothiophenes（PCDTs） are a group of important persistent organic pollutants.In the present study,geometrical optimization and electrostatic potential calculations have been performed for all 135 PCDTs congeners at the B3LYP/6-31G＊ level of theory.By means of the VSMP（variable selection and modeling based on prediction） program,one optimal descriptor（molecular polarizability,α） was selected to develop a QSRR model for the prediction of gas chromatographic retention indices（GC-RI） of PCDTs.The estimated correlation coefficients（r2） and LOO-validated correlation coefficients（q2）,all more than 0.99,were built by multiple linear regression,which shows a good estimation ability and stability of the models.A prediction power for the external samples was validated by the model built from the training set with 17 polychlorinated dibenzothiophenes.
文摘Government credibility is an important asset of contemporary national governance, an important criterion for evaluating government legitimacy, and a key factor in measuring the effectiveness of government governance. In recent years, researchers’ research on government credibility has mostly focused on exploring theories and mechanisms, with little empirical research on this topic. This article intends to apply variable selection models in the field of social statistics to the issue of government credibility, in order to achieve empirical research on government credibility and explore its core influencing factors from a statistical perspective. Specifically, this article intends to use four regression-analysis-based methods and three random-forest-based methods to study the influencing factors of government credibility in various provinces in China, and compare the performance of these seven variable selection methods in different dimensions. The research results show that there are certain differences in simplicity, accuracy, and variable importance ranking among different variable selection methods, which present different importance in the study of government credibility issues. This study provides a methodological reference for variable selection models in the field of social science research, and also offers a multidimensional comparative perspective for analyzing the influencing factors of government credibility.
文摘Ballistic missile defense system (BMDS) is important for its special role in ensuring national security and maintaining strategic balance. Research on modeling and simulation of the BMDS beforehand is essential as developing a real one requires lots of manpower and resources. BMDS is a typical complex system for its nonlinear, adaptive and uncertainty characteristics. The agent-based modeling method is well suited for the complex system whose overall behaviors are determined by interactions among individual elements. A multi-agent decision support system (DSS), which includes missile agent, radar agent and command center agent, is established based on the studies of structure and function of BMDS. Considering the constraints brought by radar, intercept missile, offensive missile and commander, the objective function of DSS is established. In order to dynamically generate the optimal interception plan, the variable neighborhood negative selection particle swarm optimization (VNNSPSO) algorithm is proposed to support the decision making of DSS. The proposed algorithm is compared with the standard PSO, constriction factor PSO (CFPSO), inertia weight linear decrease PSO (LDPSO), variable neighborhood PSO (VNPSO) algorithm from the aspects of convergence rate, iteration number, average fitness value and standard deviation. The simulation results verify the efficiency of the proposed algorithm. The multi-agent DSS is developed through the Repast simulation platform and the constructed DSS can generate intercept plans automatically and support three-dimensional dynamic display of missile defense process.
基金Sponsored by the Water Pollution Control and Management of Major Special Technology (2008ZX07317-02-03E)the Ministry of Education of Guangxi Zhuang Autonomous Region (200911MS108)Guangxi Key Laboratory of Environmental Engineering,Protection and Assessment (Nos.0801Z026,0801Z027)
文摘Using a novel freshwater photobacteria — Q67 as an indication organism and the VeritasTM luminometer with 96-well microplate as the testing equipment to determine luminous intensity of photobacteria,the familiar 29 substituted benzenes of the median inhibition toxicities（pEC50）were determined,respectively.The quantum chemical parameters of 29 substituted benzenes in the ideal gas state at 298.15 K and 1.013×105 Pa have been calculated at the B3LYP/6-31G＊ level using Gaussian 03 program.The Quantitative linear relationship（N1）between the pEC50 and two descriptors of 29 substituted benzenes was developed using the variable selection and modeling based on prediction（VSMP）.Model N1 showed good estimation ability and stability（r = 0.8777,q = 0.8482）,which exhibited the difference between empirical and predicted values of 2,3-dimethylphenol was greater（0.5）,so it was given up.Using VSMP to select the optimal descriptors,a 2-variable multiple linear regression model（called model N2）was developed for the pEC50 of substituted benzenes.The r and q for model N2 based on 28 substituted benzenes are 0.8991 and 0.8735,respectively.In order to validate the model,28 substituted benzenes were divided into a training set consisting of 20 compounds and a test set with 8 compounds.The result showed that some main structural factors influencing the pEC50 of substituted benzenes are the lowest unoccupied orbital（ELUMO）and total energy（EHF）.
文摘变量预测模型的模式识别方法(Variable predictive model based class discriminate,VPMCD)是一种利用特征值相互内在关系进行模式识别的新方法。论文提出了基于局部均值分解LMD(Local mean decomposition,LMD)能量矩概念,并针对轴承故障振动信号特征值的相互内在联系,将LMD能量矩与变量预测模型模式识别相结合,提出了一种轴承故障智能诊断新方法。首先利用LMD方法将复杂非平稳的原始信号分解为若干PF(Product function,PF)分量;然后利用相关分析剔除LMD方法中的虚假PF分量,并提取真实PF分量能量矩组成特征向量来有效地表达故障信息;最后采用VPMCD方法进行轴承故障诊断。通过仿真信号验证了PF能量矩比PF能量更能反映非平稳信号本质特征。轴承故障诊断实验结果表明,论文提出的方法能有效地应用于小样本多分类轴承故障智能诊断。
文摘提出了基于VPMCD(Variable Predictive Model Based Class Discriminate,简称VPMCD)和EMD(Empirical mode decomposition,简称EMD)的齿轮故障诊断方法,并将它应用于齿轮稳态信号的分析。VPMCD方法是一种新的模式识别方法,特别适合于非线性分类问题,它充分利用从原始数据中所提取的特征值之间的相互内在关系建立数学模型,从而进行模式识别。在基于VPMCD和EMD的齿轮故障诊断方法中,首先采用EMD方法将齿轮振动信号自适应地分解为若干个单分量信号,然后提取各个分量的样本熵并将其作为特征值,最后采用VPMCD分类器进行故障识别和分类。结果表明该方法能够有效地突出齿轮故障振动信号的故障特征,提高了齿轮故障诊断的准确性。
文摘针对滚动轴承故障诊断时所提取的特征值中可能含有较小相关性和冗余性特征,采用基于Wrapper模式的距离评价技术(distance evaluation technique,简称DET)进行特征选择。在分类器的设计中,提出了基于稳健回归的多变量预测模型(Robust regression-Variable predictive model based class discriminate,简称RRVPMCD)分类方法,以减小"异常值"对参数估计的影响,从而有望建立更加准确的预测模型。即根据Wrapper模式的特点,首先通过DET方法计算出各特征值对类的敏感度,并结合RRVPMCD分类器,选择敏感度最大的若干特征值组成特征向量矩阵;然后用RRVPMCD方法进行训练,建立预测模型;最后用所建立的预测模型进行模式识别。实验分析结果表明,基于Wrapper模式的特征选择方法和RRVPMCD分类方法相结合可以有效地对滚动轴承的工作状态和故障类型进行识别。
文摘提出一种基于拉普拉斯特征映射流形学习算法(Laplacian Eigenmaps,简称LE)和改进多变量预测模型(Variable predictive model based class discriminate,简称VPMCD)的滚动轴承故障诊断方法,首先对振动信号进行局部特征尺度分解(Local characteristic scale decomposition,简称LCD),并提取各内禀尺度分量(Intrinsic scale component,简称ISC)的特征构造高维特征向量,接着采用LE算法挖掘出高维数据中包含有效信息且具有内在规律性的低维特征,然后输入到基于Kriging的改进多变量预测模型(Kriging-variable predictive model based class discriminate,简称KVPMCD)分类器中进行模式识别。该方法充分利用并有效结合了LCD在信号处理、LE在挖掘特征信息和KVPMCD在模式识别方面的优势,实现了滚动轴承故障特征提取到故障识别的全程诊断。实验分析结果表明:基于LE算法和KVPMCD的分类方法可以有效地对滚动轴承的工作状态和故障类型进行识别。
文摘VPMCD(Variable Predictive Model Based Class Discriminate)是一种新的模式识别方法,它充分利用从原始数据中所提取的特征值之间的相互内在关系建立数学模型,从而进行模式识别。论文将VPMCD结合排列熵(Permutation Entropy,简称PE)方法应用于滚动轴承故障诊断。首先采用ITD(Intrinsic Time-scale Decomposition,简称ITD)对滚动轴承振动信号进行分解,得到若干个固有旋转(Proper Rotation)分量,并对包含主要故障信息的PR分量提取排列熵作为故障特征值;然后,对VPMCD分类器进行训练;最后,采用VPMCD分类器进行故障识别和分类。实验数据的分析结果表明该方法能够有效地应用于滚动轴承故障诊断。
文摘基于变量预测模型的分类识别(Variable predictive model-based class discriminate,VPMCD)方法是一种新的分类识别方法,但模型类型的选择存在主观性。为了解决VPMCD方法应用于机械故障诊断过程中的模型选择问题,结合遗传算法的全局优化能力,提出了基于GA-VPMCD(Genetic algorithm and variable predictive model based class discriminate)智能诊断方法。首先通过样本训练建立多个弱VPM(Variable predictive model),然后采用遗传算法优化各个弱VPM的权值,得到最优权值矩阵,最后用最优权值矩阵加权融合测试样本的弱VPM特征变量预测值,得到最佳特征变量预测值,并以误差平方和最小为辨别函数分类识别故障类型。通过GA-VPMCD方法在滚动轴承故障智能诊断中的应用实验验证了基于GA-VPMCD的故障智能诊断方法能有效地提高诊断精度和诊断系统的鲁棒性。
文摘基于变量预测模型的模式识别(variable predictive model based class discriminate,VPMCD)方法是一种充分利用特征值之间相互内在关系进行多分类模式识别的新方法。对VPMCD算法进行了研究,并采用交叉验证法来选择VPMCD模型。针对机械故障振动信号的特征值之间的相互内在关系,结合本征时间尺度分解(intrinsic time-scale decom-position,ITD),提出了一种基于本征时间尺度分解和VPMCD的机械故障诊断方法。该方法首先利用ITD方法将原始信号分解若干个PR(proper rotation,PR)分量,然后提取第一个PR分量的无量纲时域统计参数组成特征向量,最后采用VPMCD方法进行机械故障诊断。通过滚动轴承故障诊断实验验证了该方法能有效地应用于小样本多分类机械故障诊断。
文摘将基于变量预测模型的模式识别(variable predictive model based class discriminate,简称VPMCD)、独立分量分析(independent component analysis,简称ICA)和相关系数分析方法相结合,提出了基于ICA相关系数和VPMCD的滚动轴承故障诊断方法。首先,对不同工况下的滚动轴承振动信号分别进行独立分量分析,获得各工况信号的独立分量;然后,提取样本与不同工况信号独立分量之间的相关系数,并以相关系数绝对值的和作为该样本的特征值;最后,采用VPMCD分类器进行故障识别和分类。实验数据的分析结果表明,该方法能够有效应用于滚动轴承故障诊断。
