欢迎您！东篱公司
退出
-
申报数据库
-
立项数据库
-
成果数据库
-
项目获奖数据库
中文摘要：
针对多变量预测模型（VPMCD）模式识别方法的固有缺陷和机械故障特征难以选择的难题,即特征维数较多时对时效性的影响和特征选择需要引入主观因素的现状,提出了一种基于嵌入式的逐步回归多变量预测模型（SRVPMCD）模式识别方法。该方法首先通过逐步回归引入变量并计算其显著水平,建立只包含显著特征值的预测模型,同时实现嵌入式特征选择和建模分类的功能,然后用所建立的预测模型来预测待分类样本的特征值,最后把预测结果作为分类依据进行模式识别。对滚动轴承故障信号的分析结果表明,基于嵌入式SRVPMCD的模式识别方法可以实现特征选择和分类的双重功能,在保证识别精度的前提下,比原VPMCD方法及其组合方法可以更快地识别滚动轴承的工作状态和故障类型。
英文摘要：
Targeting the inherent defects of the variable predictive model based class discriminate（VPMCD）and the problems of how to choose the features of those mechanical faults,which was the effects on timeliness due to the excessive characteristic dimensions and feature selection needed o introduce the subjective factors,a new embedded SRVPMCD method was put forward herein.Introducing variable and calculating the significant level by stepwise regression,establishing aprediction model that only contained significant characteristic values,achieving the function of embedded feature selection and modeling simultaneously,and then using the established prediction model to predict the characteristic values of those unclassified signals samples,the prediction results would be recognized by the model as accordance to classify.The analysis results for roller bearing fault signals show that compared to the original VPMCD and combined methods,the pattern recognition method based on embedded SRVPMCD can realize the double functions of feature selection and classification,and identify the working states and fault types of roller bearing quickly.
同期刊论文项目
同项目期刊论文
Rolling bearing fault diagnosis based on multi-scale fuzzy entropy and variable predictive model-bas
Normalized complex Teager energy operator demodulation method and its application to fault diagnosis
Generalized empirical mode decomposition and its applications to rolling element bearing fault diagn
期刊信息
位置：