Dictionary Learning via a Mixed Noise Model for Sparse Representation Classification of Rolling Bearings

Abstract

Rotating machinery contains a great number of rolling bearings, which play an indispensable role. However, bearing vibration signals in complex environments are often mixed with various noises, which makes it difficult to extract fault characteristics from original signals. It is still challenging to identify the fault types of rolling bearings. To address this issue, a dictionary learning method based on a mixed noise model for the sparse representation classification of rolling bearings (DLMN-SRC) is proposed. Our framework constructs a dictionary learning method based on mixed noise, which has better robustness to complex noise pollution than a single noise model. Then, an alternating direction method of multipliers (ADMM) algorithm is used to solve the optimization problem of the proposed model. Eventually, the redundant errors between the detected and reconstructed signals in the dictionary learning model are calculated for sparse representation classification. The results of two examples prove that faults of the rolling bearing are successfully extracted and classified by DLMN-SRC. Compared with traditional diagnosis methods, the performance of this method has obvious superiority and good application prospects.

A weighted sparse classification method based on period analysis dictionary

“…Learning dictionaries have been widely used in dictionary construction for their adaptive properties. Zhang et al [19] developed a sparse classification method based on a noise model using a K-SVD training dictionary. Kong et al [20] introduced an overlapping segmentation strategy to enhance the training dataset and proposed a sparse learning-based classification framework.…”

Fault Feature Extraction Method for Rolling Bearings Based on Complete Ensemble Empirical Mode Decomposition with Adaptive Noise and Variational Mode Decomposition

“…As one of the main parts of rotating machinery [1][2][3][4], rolling bearings are widely used in high-end manufacturing, including aviation [5,6], automobiles [7,8], and ships [9,10]. However, influenced by load, installation, and lubrication factors, a variety of failures appear [11,12], which are easily overwhelmed by complex noise in a complex and changeable working environment. Thus, effective fault features cannot be directly extracted [13][14][15].…”