Rotating machinery fault diagnosis based on impact feature extraction deep neural network

Abstract

Gears and bearings are important components in rotating machinery and are crucial for the safety and operation of the whole mechanical system. Intelligent fault diagnosis methods based on deep-learning algorithms have undergone rapid development in recent years. Despite this, integrating fault features in a deep network construction remains a challenge for intelligent fault diagnosis of rotating machinery. In this paper, a novel impact feature extraction deep neural network (IFE-DN) is proposed for intelligent gear and bearing fault diagnosis. An improved three-layer Laplace wavelet kernel convolutional neural network (LW-CNN), where the Laplace wavelet kernel is designed in the first convolutional layer, is constructed to extract and enhance the impact features in the vibration signal. Using a visualized heat map, the physical meaning of the LW-CNN’s extracted features is explained and the interpretability of the network model is enhanced. The wavelet function selection in the deep neural network is also discussed. The extracted features are transferred to a primary capsule layer and a digital capsule layer. With a feature vector converting process and dynamic routing algorithm, more detailed features are optimized and the fault types are classified. Four experimental data sets from different laboratories are used to verify the performance of the proposed model, and t-distributed stochastic neighbour embedding is carried out to visually analyze the extracted features in different layers. The results of the analysis of gear and bearing faults of different types and defect sizes show that the IFE-DN presents significant accuracy and satisfactory generalization ability.

A novel multi-featured decision system for multi-classification tasks

“…At the feature selection level, dimensionality reduction algorithms in machine learning are the primary strategy to consider because of the huge impact of redundant features on signal quality. Principal component analysis [15], linear discriminant analysis (LDA), isometric mapping [16], locally linear embedding [17,18] and Laplacian eigenmaps [19], multi-dimensional scaling [20], t-distributed stochastic neighbor embedding [21,22] are classical dimensionality reduction algorithms, and these can extract the essential features of the signals. But on the issue of signal classification, it is necessary to pay attention not only to the essential characteristics of signals, but also to the common characteristics of the same class of signals and the individual characteristics among different classes of signals, and this guidance is reflected in the need to design feature selection algorithms based on known class labels.…”

Federated learning for intelligent fault diagnosis based on similarity collaboration

“…Machinery health monitoring plays an important role in the field of fault diagnosis, reducing production costs, optimizing maintenance programs, and increasing productivity [1][2][3]. Since the development of deep learning techniques, datadriven methods have proven effective in machinery health monitoring and have been widely applied in many fields, such as manufacturing, energy, and transportation.…”

DEC-NWD: an approach for open set domain adaptation in fault diagnosis

“…Rolling bearings play a crucial role in rotating machinery like wind power generation equipment, high-precision CNC machine tools, and aerospace engines [1]. However, prolonged operation under harsh conditions can lead to bearing damage, which in turn can result in unexpected machine malfunctions.…”