对于滚动轴承的智能故障诊断问题,文章提出一种基于梯度直方图(histogram of oriented gradient,HOG)图像处理的轴承故障诊断方法.首先将传感器采集到的原始时域振动信号经变分模态分解(variation-al mode decomposition,VMD)后得到二维图像;再利用HOG数字图像处理算法对上述图像提取特征信息;使用多维尺度分析(multi-dimensional scaling,MDS)方法对特征数据进行降维处理,得到低维空间下的故障特征数据并对故障特征数据添加标签构建数据集,该数据集被划分为训练集和测试集;然后引入遗传算法(ge-netic algorithm,GA)优化支持向量机(support vector machine,SVM)中的惩罚因子和核函数关键参数,用训练集进行训练得到最优故障分类模型;最后对测试集的数据进行处理,得到分类结果.对比分析结果表明,该文方法能快速提取轴承故障有效特征,提高故障诊断准确率.
Rolling bearing fault diagnosis method based on HOG image processing
For the problem of intelligent fault diagnosis of rolling bearings,a method of bearing fault i-dentification and diagnosis based on histogram of oriented gradient(HOG)image processing algorithm is proposed.Firstly,the original time-domain vibration signal collected by the sensor is subjected to variational mode decomposition(VMD)to obtain a two-dimensional image.The HOG digital image processing algorithm is used to extract the feature information of the above image.Multi-dimensional scaling(MDS)analysis is used to reduce the dimensionality of the feature data to obtain the fault fea-ture data in the low-dimensional space,and labels are added to the fault feature data to construct a da-ta set.The data set is divided into a training set and a test set.Then,the genetic algorithm(GA)is introduced to optimize the penalty factor and the key parameters of the kernel function in the support vector machine(SVM),and the training set is used for training to obtain the optimal fault classifica-tion model.Finally,the data of the test set are processed to obtain the classification result.The com-parative analysis results show that the method in this paper can quickly extract the effective features of bearing faults and improve the accuracy of fault diagnosis.
histogram of oriented gradient(HOG)image processingfault diagnosismulti-dimen-sional scaling(MDS)rolling bearing
万方数据
维普
