According to statistic data,machinery faults contribute to largest proportion of High-voltage circuit breaker failures,and traditional maintenance methods exist some disadvantages for that issue.Therefore,based on the...According to statistic data,machinery faults contribute to largest proportion of High-voltage circuit breaker failures,and traditional maintenance methods exist some disadvantages for that issue.Therefore,based on the wavelet packet decomposition approach and support vector machines,a new diagnosis model is proposed for such fault diagnoses in this study.The vibration eigenvalue extraction is analyzed through wavelet packet decomposition,and a four-layer support vector machine is constituted as a fault classifier.The Gaussian radial basis function is employed as the kernel function for the classifier.The penalty parameter c and kernel parameterδof the support vector machine are vital for the diagnostic accuracy,and these parameters must be carefully predetermined.Thus,a particle swarm optimizationsupport vector machine model is developed in which the optimal parameters c andδfor the support vector machine in each layer are determined by the particle swarm algorithm.The validity of this fault diagnosis model is determined with a real dataset from the operation experiment.Moreover,comparative investigations of fault diagnosis experiments with a normal support vector machine and a particle swarm optimization back-propagation neural network are also implemented.The results indicate that the proposed fault diagnosis model yields better accuracy and e-ciency than these other models.展开更多
提出了一种新的支持向量机(Support V ectorM ach ines,SVM)机械系统状态组合预测模型。应用FPE(F ina lP rinc ip le E rror)准则优化样本的维数,采用时域内的振动烈度和频域内的特征频率分量作为预测机械系统状态的敏感因子,构建了预...提出了一种新的支持向量机(Support V ectorM ach ines,SVM)机械系统状态组合预测模型。应用FPE(F ina lP rinc ip le E rror)准则优化样本的维数,采用时域内的振动烈度和频域内的特征频率分量作为预测机械系统状态的敏感因子,构建了预测模型。支持向量机采用新型的结构风险最优化准则,预测能力强、鲁棒性好。采用径向基函数和ε损失函数,将该模型应用于实验台和旋转注水机组的状态预测,取得了较好的效果。这表明利用支持向量机的组合预测模型,可以降低设备维修代价,提高设备的安全性和可靠性。展开更多
机械振动信号携带大量重要的机械状态信息,然而机械故障振动信号在复杂工作状态下通常呈现非平稳、非线性特性。因此,从振动信号抽取和选择有效的机械故障特征、提高故障识别性能,成为机械故障诊断研究的热点。针对上述问题,本文提出了...机械振动信号携带大量重要的机械状态信息,然而机械故障振动信号在复杂工作状态下通常呈现非平稳、非线性特性。因此,从振动信号抽取和选择有效的机械故障特征、提高故障识别性能,成为机械故障诊断研究的热点。针对上述问题,本文提出了基于集成局部均值分解(Ensemble local means decomposition,ELMD)与改进的稀疏多尺度支持向量机(Sparse multiscale support vector machine,SMSVM)的机械故障诊断方法。该方法首先使用自适应非线性、非平稳信号处理方法 ELMD把多模态调制故障信号分解成为多个单模态解调信号,有效地增强了故障特征。把压缩感知和多尺度分析技术融合于故障模式分类中,提出改进SMSVM旋转机械故障识别方法,提高多类机械微弱故障数据模式识别性能。该方法融合稀疏表示、多尺度分析和SVM的优点,无需求解复杂的优化问题,易于推广至更多尺度SVM,具有计算量少、泛化性与鲁棒性好、物理意义明显等优点。人工数据和实验设备数据验证了本文算法的优越性。展开更多
提出基于Laplacian双联最小二乘支持向量机(Laplacian Twin Least Squares Support Vector Machine,LapTLSSVM)半监督模式识别的新型早期故障诊断方法。用时、频域特征集广泛收集旋转机械不同早期故障的特征信息,再用提升半监督局部Fis...提出基于Laplacian双联最小二乘支持向量机(Laplacian Twin Least Squares Support Vector Machine,LapTLSSVM)半监督模式识别的新型早期故障诊断方法。用时、频域特征集广泛收集旋转机械不同早期故障的特征信息,再用提升半监督局部Fisher判别分析(Enhanced Semi-Supervised Local Fisher Discriminant Analysis,ESSLFDA)将高维时、频域特征集约简为具有更好类区分度的低维特征向量,并输入到Lap-TLSSVM中进行早期故障诊断。Lap-TLSSVM引入了包含大量无标签数据信息的流形规则实现半监督学习;其目标函数只含等式约束条件,且用共轭梯度法求解目标函数的线性方程组以加速训练过程。所提出的方法在训练样本非常稀少的情况下具有较高的诊断精度和计算效率。深沟球轴承早期故障诊断实验验证了该方法的有效性。展开更多
基金Supported by National Natural Science Foundation of China(Grant No.51705372)National Science and Technology Project of the Power Grid of China(Grant No.5211DS16002L).
文摘According to statistic data,machinery faults contribute to largest proportion of High-voltage circuit breaker failures,and traditional maintenance methods exist some disadvantages for that issue.Therefore,based on the wavelet packet decomposition approach and support vector machines,a new diagnosis model is proposed for such fault diagnoses in this study.The vibration eigenvalue extraction is analyzed through wavelet packet decomposition,and a four-layer support vector machine is constituted as a fault classifier.The Gaussian radial basis function is employed as the kernel function for the classifier.The penalty parameter c and kernel parameterδof the support vector machine are vital for the diagnostic accuracy,and these parameters must be carefully predetermined.Thus,a particle swarm optimizationsupport vector machine model is developed in which the optimal parameters c andδfor the support vector machine in each layer are determined by the particle swarm algorithm.The validity of this fault diagnosis model is determined with a real dataset from the operation experiment.Moreover,comparative investigations of fault diagnosis experiments with a normal support vector machine and a particle swarm optimization back-propagation neural network are also implemented.The results indicate that the proposed fault diagnosis model yields better accuracy and e-ciency than these other models.
文摘提出了一种新的支持向量机(Support V ectorM ach ines,SVM)机械系统状态组合预测模型。应用FPE(F ina lP rinc ip le E rror)准则优化样本的维数,采用时域内的振动烈度和频域内的特征频率分量作为预测机械系统状态的敏感因子,构建了预测模型。支持向量机采用新型的结构风险最优化准则,预测能力强、鲁棒性好。采用径向基函数和ε损失函数,将该模型应用于实验台和旋转注水机组的状态预测,取得了较好的效果。这表明利用支持向量机的组合预测模型,可以降低设备维修代价,提高设备的安全性和可靠性。
文摘机械振动信号携带大量重要的机械状态信息,然而机械故障振动信号在复杂工作状态下通常呈现非平稳、非线性特性。因此,从振动信号抽取和选择有效的机械故障特征、提高故障识别性能,成为机械故障诊断研究的热点。针对上述问题,本文提出了基于集成局部均值分解(Ensemble local means decomposition,ELMD)与改进的稀疏多尺度支持向量机(Sparse multiscale support vector machine,SMSVM)的机械故障诊断方法。该方法首先使用自适应非线性、非平稳信号处理方法 ELMD把多模态调制故障信号分解成为多个单模态解调信号,有效地增强了故障特征。把压缩感知和多尺度分析技术融合于故障模式分类中,提出改进SMSVM旋转机械故障识别方法,提高多类机械微弱故障数据模式识别性能。该方法融合稀疏表示、多尺度分析和SVM的优点,无需求解复杂的优化问题,易于推广至更多尺度SVM,具有计算量少、泛化性与鲁棒性好、物理意义明显等优点。人工数据和实验设备数据验证了本文算法的优越性。
文摘提出基于Laplacian双联最小二乘支持向量机(Laplacian Twin Least Squares Support Vector Machine,LapTLSSVM)半监督模式识别的新型早期故障诊断方法。用时、频域特征集广泛收集旋转机械不同早期故障的特征信息,再用提升半监督局部Fisher判别分析(Enhanced Semi-Supervised Local Fisher Discriminant Analysis,ESSLFDA)将高维时、频域特征集约简为具有更好类区分度的低维特征向量,并输入到Lap-TLSSVM中进行早期故障诊断。Lap-TLSSVM引入了包含大量无标签数据信息的流形规则实现半监督学习;其目标函数只含等式约束条件,且用共轭梯度法求解目标函数的线性方程组以加速训练过程。所提出的方法在训练样本非常稀少的情况下具有较高的诊断精度和计算效率。深沟球轴承早期故障诊断实验验证了该方法的有效性。
