Evaluation Of Health Status Of Rolling Bearings Based On Adaptive Modal Decomposition

With the continuous progress of human civilization and the development of science and technology,higher standards are put forward for the reliability,safety and economy of equipment operation.Health state assessment is an important basic technology which uses the monitored signal data to assess the current equipment state.The application of this technology can effectively reduce the failure rate,realize regular maintenance and repair,and ensure the normality of mechanical equipment operation to avoid major accidents.Rolling bearings are key components of the mechanical equipments,and their operation situations directly affect the performance of the whole machines.Therefore,rolling bearings are selected as the research objects of this paper.The research content of this paper mainly includes:（1）Research on feature extraction based on adaptive mode decompositionAdaptive mode decomposition algorithm has the characteristics of complete data driving and self-adaptability in data analysis.Therefore,this paper uses adaptive mode decomposition algorithm to extract the characteristics of the non-linear and non-stationary rolling bearing vibration signal throughout the life cycle.This paper selects three adaptive decomposition algorithms,empirical mode decomposition（EMD）,ensemble empirical mode decomposition（EEMD）and variational mode decomposition（VMD）to decompose the original vibration signal and obtain a series of intrinsic mode functions（IMF）.The IMFs with high information relevance to the original signal are screened out,and the time-domain features of the IMF are extracted respectively to construct the initial multi-dimensional feature space.（2）Research on the improved Mahalanobis-Taguchi System（MTS）based on rough set and random forest weightingThe MTS is an important classification method.Because of the shortcomings of the traditional MTS,this paper use rough set method to replace the orthogonal table and the signal-to-noise ratio to screen the feature variables.After this kind of improvement,the dimensionality of the feature is reduced and the reduced feature space is obtained.Then the optimized feature parameters are input into the MTS and random forest is used to assign weights to different feature parameters to obtain the weighted Mahalanobis distance which lay the foundation for follow-up research.（3）Health state assessment method based on adaptive mode decomposition and improved Mahalanobis-Taguchi systemCombining three adaptive decomposition algorithms with MTS method,EMD-improved MTS based health assessment method,EEMD-improved MTS based health assessment method,and VMD-improved MTS,this paper constructs three health state evaluation models by three methods,and then uses the arctangent function to construct the health index to realize the health state assessment of rolling bearings.（4）Health state assessment method based on fusion modelIn order to fully reflect the health state of rolling bearings,the Sigmoid function is used to construct a health state fusion model and to achieve the fusion of decision-making levels.This paper uses the data from Rolling Bearing Experimental Platform in the University of Cincinnati to evaluate the health of rolling bearings.The empirical results show that the information fusion method can greatly improve the accuracy and reliability of the health assessment of rolling bearings,and can better identify early degraded states.