A Novel Study on a Generalized Model Based on Self-Supervised Learning and Sparse Filtering for Intelligent Bearing Fault Diagnosis

Abstract

Recently, deep learning has become more and more extensive in the field of fault diagnosis. However, most deep learning methods rely on large amounts of labeled data to train the model, which leads to their poor generalized ability in the application of different scenarios. To overcome this deficiency, this paper proposes a novel generalized model based on self-supervised learning and sparse filtering (GSLSF). The proposed method includes two stages. Firstly (1), considering the representation of samples on fault and working condition information, designing self-supervised learning pretext tasks and pseudo-labels, and establishing a pre-trained model based on sparse filtering. Secondly (2), a knowledge transfer mechanism from the pre-training model to the target task is established, the fault features of the deep representation are extracted based on the sparse filtering model, and softmax regression is applied to distinguish the type of failure. This method can observably enhance the model’s diagnostic performance and generalization ability with limited training data. The validity of the method is proved by the fault diagnosis results of two bearing datasets.

A new model for bearing fault diagnosis based on mutual mapping of signals and images and sparse representation

“…As shown in table 14, we have selected some existing models from the past 5 years to compare with the models in this paper. When using the same number of samples as [26], except for slightly lower accuracy under 0HP data, the accuracy under the other three single working conditions is 3.48%, 3.78%, and 0.8% higher than reference, respectively. In addition, when achieving the same level of accuracy as other literature models, the amount of data used in this paper is significantly smaller.…”

A Universal Feature Extractor Based on Self-Supervised Pre-Training for Fault Diagnosis of Rotating Machinery under Limited Data

“…The main self-supervised task of their architecture is to predict future noise based on past noise. Nie et al [28] proposed a generalized model based on self-supervised learning and sparse filtering, which employs corresponding labels assigned to signals undergoing different feature transformations for self-supervised learning.…”

“…In addition, the model architectures employed exhibited several shortcomings, posing challenges to achieving high diagnostic accuracy in complex real-world settings for fault diagnosis purposes. Specifically, Wang and Nie's method [22,28] only identifies data augmentation categories without instance-level representation learning, limiting its ability to extract robust fault features effectively. In contrast, for contrastive learning methods, Ding's method [23] employs a time-domain transformation for vibration signal data augmentation but fails to leverage the time-and frequency-domain characteristics simultaneously.…”