Prediction of Wind Turbine Blades Icing Based on CJBM With Imbalanced Data

Due to the coupling of multiple fault feature information and contamination of heavy background noise, it is a challenging task to accurately identify Rolling Bearing Compound Fault (RBCF). A method for isolating and identifying the RBCF is proposed by integrating the Adaptive Periodized Singular Spectrum Analysis (APSSA) with Rényi entropy (RE). The adaptive selection of embedding dimension of Hankel matrix in APSSA without setting parameters empirically is proposed, and a selection criterion for singular values is established to preprocess the vibration signals of rolling bearing and enhance fault periodic component. A RE-based threshold value is introduced to further isolate and decouple the impulse segments of vibration signal in the time domain. Considering inner raceway fault, outer raceway fault, ball fault and skidding, a comprehensive simulation model of compound fault is constructed by the response mechanism of different excited resources. Simulated and experimental data are applied to validate the effectiveness and practicability of proposed method. The results demonstrate that RBCF can be identified correctly by the proposed method under strong background noise.

“…The elements of the new matrix y ij * = y ij if L < M, otherwise y ij * = y ji . The matrix Y is transformed into the series g m , which is calculated as equation (5):…”

Section: Singular Spectrum Analysismentioning

confidence: 99%

Isolation and identification of rolling bearing compound faults based on adaptive periodized singular spectrum analysis and Rényi entropy

Li,

Yan,

Hou

et al. 2024

“…The second category focuses on integrating conventional machine learning algorithms for classification of bearing signals in various health status. Techniques like support vector machine [9], principal component analysis [10], boosting machine [11] and random forest [12] are employed in this regard. Deep learning-based fault diagnosis methods involve extracting deep feature representations from bearing signals using neural network models like convolutional neural network (CNN) [13], AE [14], and other network architectures.…”

Section: Introductionmentioning

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.

“…activities. Data-driven fault diagnosis methods demonstrate excellent anomaly detection and prediction capabilities, particularly in the monitoring of complex equipment states (such as wind turbines) [2] and the detection of core components' states in industrial equipment (such as bearings, gearboxes, and inverters) [3,4], exhibiting more comprehensive and accurate diagnostic performance. In recent years, there has been increasing attention to anomaly sound detection (ASD) systems [5].…”

Section: Introductionmentioning

confidence: 99%

Anomaly sound detection of industrial devices by using teacher-student incremental continual learning

Zhou,

Wang,

et al. 2024

Normal production processes will be substantially impacted by industrial devices in abnormal working conditions. Anomaly sound detection (ASD) model can monitor the working condition of devices by the non-contact and non-invasive way. When new device data is introduced, traditional ASD models are trained using data from all devices, to accommodate every device. However, in real-world settings, the kinds and amounts of devices are constantly changing, which raises difficulties for the current ASD models. This paper proposes a teacher-student incremental learning method for ASD models, aiming to solve ASD model scalability problem. In this paradigm, teacher model has knowledge of all the old devices. The objective of student model is to learn new device knowledge, while avoiding the forgetting of old device knowledge. When student model learns new device data, teacher model transfers the acoustic feature knowledge of old devices to student model via knowledge distillation. Furthermore, the imbalance between old and new knowledge causes challenges, such as knowledge forgetting or lower learning efficiency for student model. This paper presents a dual-teacher-student (DTS) model to solve the problem of knowledge imbalance. Different teacher models for new and old devices in DTS, directing student model to accomplish continuous and deep integration of knowledge. Evaluation for proposed method on the DCASE 2020 Task2 dataset. The results show, the proposed method outperforms other methods in terms of learning capability and robustness during the incremental learning process. Analysis of significance test on the experimental results demonstrates that the method outperforms other methods statistically.