A novel diagnosis methodology of gear oil for wind turbine combining stepwise multivariate regression and clustered federated learning framework

Research Square (Research Square), Journal Year: 2025, Volume and Issue: unknown

Published: April 23, 2025

Data-driven approaches demonstrate significant potential in accurately diagnosing faults wind turbines. To enhance diagnostic performance, we introduce a clustered federated learning framework (CFLF) to gear oil diagnosis. Initially, stepwise multivariate regression (SMR) model is introduced and optimized after data process, which integrates multiscale feature AIC diagnosis feature. Subsequently, tackle heterogeneity among different indicators, canonical correlation series of representations are extracted from the SMR models, combining CFLF method proposed assess performance oil. Actual analysis turbine showcase superior over single with higher prediction accuracy 35.73%. This study provides new technique for evaluating energy sector.

Language: Английский

Vibration signal analysis for rolling bearings faults diagnosis based on deep-shallow features fusion

Scientific Reports, Journal Year: 2025, Volume and Issue: 15(1)

Published: March 18, 2025

Abstract In engineering applications, the bearing faults diagnosis is essential for maintaining reliability and extending lifespan of rotating machinery, thereby preventing unexpected industrial production downtime. Prompt fault using vibration signals vital to ensure seamless operation system avert catastrophic breakdowns, reduce maintenance costs, continuous productivity. As industries evolve machines operate under diverse conditions, traditional detection methods often fall short. spite significant research in recent years, there remains a pressing need improve existing diagnosis. To fill this gap, work aims propose an efficient robust diagnosing faults, deep Shallow features. Through evaluated experiments, our proposed model Multi-Block Histograms Local Phase Quantization (MBH-LPQ) showed excellent performance classification accuracy, audio-trained VGGish best all tasks. Contributions include: Combine descriptor, derived from novel hand-crafted discriminative features MBH-LPQ, with obtained pre-trained Convolutional Neural Network (CNN) audio spectrograms, by merging at score level Weighted Sum (WS). This approach designed take advantage complementary strengths both feature models, thus enhancing overall diagnostic performance. Furthermore, experiments conducted verify approach’s assessed based on accuracy demonstrated rate two different noisy datasets, 98.95% 100% being reached CWRU PU datasets benchmark, respectively.

Language: Английский

A hybrid approach combining deep learning and signal processing for bearing fault diagnosis under imbalanced samples and multiple operating conditions

Scientific Reports, Journal Year: 2025, Volume and Issue: 15(1)

Published: April 19, 2025

To enhance bearing fault diagnosis performance under various operating conditions, this paper proposes a hybrid approach based on generative adversarial networks (GANs), transfer learning, wavelet transform time-frequency representations, asymmetric convolutional networks, and the multi-head attention mechanism (MAC-MHA). Firstly, GANs are utilized to generate new data meet model's training requirements. Then, is applied convert vibration signals into capturing temporal evolution of frequency components. Next, an improved network (MAC-MHA), combined with mechanism, employed focus key features, further improving accuracy. Considering differences in learning techniques facilitate knowledge from source domain target domain, thereby enhancing generalization ability. Experimental results demonstrate effectiveness robustness proposed method conditions. Finally, validated using PADERBORN CWRU datasets.

Language: Английский