CNN parameter design based on fault signal analysis and its application in bearing fault diagnosis

Fault diagnosis of rolling bearings in complex environments is a difficult problem. First, the median filter can remove the noise in the vibration signals, however, it cannot adaptively adjust the filter weights according to the input signals. Second, the popular vision transformer (ViT) cannot extract local feature information under complex conditions and has a large number of parameters, which result in increased computational complexity. To solve these problems, a lightweight multi-feature fusion ViT bearing fault diagnosis method with strong local awareness in complex environments is proposed. Firstly, to learn the features and statistical distributions of the input signals, the gradient descent method is used to continuously and iteratively update the weights and filter the signals. Then, to better extract critical local fault information, a local sensing module is constructed using MWCNN. Finally, an improved lightweight multi-feature fusion ViT is constructed to perform global feature extraction and fault identification. The results show that the proposed method has better noise reduction effect and feature extraction ability, and can accurately identify the fault types under the complex environments.

Section: Model Parameters and Comparison Methodsmentioning

confidence: 99%

A lightweight multi-feature fusion vision transformer bearing fault diagnosis method with strong local sensing ability in complex environments

Li,

Zhao

2024

“…(2) Concatenating normal bearing data with nine distinct fault datasets, the model is continuously fed with data stream batches to evaluate its online diagnosis ability. Since EMD and CNN are widely used in the field of fault diagnosis of bearing data [29,30], and the proposed method uses SLDA to achieve online fault diagnosis, in order to verify the effectiveness of the proposed method, algorithm comparison is conducted using the fault diagnosis method that combines EMD and CNN, as well as SAX and SDCAE algorithm. The approach, using EMD and CNN, decomposes the vibration signal into intrinsic mode functions (IMFs) through EMD, followed by Fast Fourier Transform on maximum IMF, subsequently normalizing it to acquire the frequency domain signal.…”

Section: Case Western Reserve University (Cwru) Bearing Data Casementioning

confidence: 99%

Bearing fault diagnosis based on online symbolic aggregation approximation and streaming deep discriminant analysis

Wang,

Ye,

Zeng

2024

Rolling bearings are one of the important components of many industrial equipment, and the timely detection of faults in these bearings significantly contributes to ensuring equipment safety. To achieve real-time diagnosis of bearing faults, this paper proposes an online fault diagnosis framework that utilizes online symbolic aggregation approximation (SAX) and streaming deep discriminant analysis. This method uses the moving window approach to segment the vibration data collected online, and then uses SAX for symbolic representation.The obtained icons are input into stacked denoising convolutional autoencoder (SDCAE) forclassification. The model consists of several denoising convolutional autoencoders (DCAEs) and a linear discriminant analysis (LDA) module. To accommodate online data changes, while fixing the main structure of the network, the streaming method is used to update the parameters of LDA. Compared with other traditional approaches for diagnosing bearing faults, the proposed method exhibits distinct advantages. Specifically, it utilizes a stable neural network model, updates classifier parameters in real-time, and demonstrates superior computational efficiency. The application results demonstrate the effectiveness of the proposed method.

“…With the continuous development of artificial intelligence technology, deep learning (DL) has been introduced into the field of bearing fault diagnosis by more and more scholars [8], such as convolutional neural network (CNN) [9], recurrent neural network [10], graph neural network [11], and other networks have achieved remarkable success through their intelligent processing of large amounts of data, automatic extraction and learning of discriminative features, and high discrimination. However, DL relies on many labeled and co-distributed data during model training, which is unachievable in real industrial scenarios.…”

Section: Introductionmentioning

confidence: 99%

Intelligent fault diagnosis of double-aligned domain adaptation based on multi-structure fusion and multi-attention mechanism

Shang,

Wu,

Pan

et al. 2024

Currently, most fault diagnosis methods based on domain adaptation reduce the distribution discrepancy between the source and target domains by aligning their high-dimensional features, but do not consider the impact of the source domain decision boundary on the distribution of the target domain, which leads to slow convergence and poor generalization. Aiming at the above-mentioned issues, this paper proposes a double-aligned domain adaptation deep transfer model (MSADA) based on multi-structure fusion and multi-attention mechanisms. Firstly, a multi-structure fusion network with Inception module is constructed to increase the information capacity in the extracted features and enhance the representation of deep features. Then, the multi-attention mechanism is embedded into different convolutional channels separately to learn globally and locally important information and reduce the influence of the source domain decision boundary on the target domain distribution. Finally, the multi-kernel maximum mean discrepancy (MK-MMD) and KL divergence are combined as a new double-aligned distribution discrepancy metric to align the samples and probability distributions of the source domain and the target domain, respectively. At the same time, a dynamic adaptive factor is designed to adjust the contribution of the two types of distributions, thus effectively improving the training efficiency and the robustness of the model. Through the validation analysis of two rolling bearing dataset cases, the proposed MSADA has better cross-domain diagnostic performance than other domain adaptation methods.