An adaptive hierarchical hybrid kernel ELM optimized by aquila optimizer algorithm for bearing fault diagnosis DOI Creative Commons
Hao Yan, Liangliang Shang,

Wan Chen

и другие.

Scientific Reports, Год журнала: 2025, Номер 15(1)

Опубликована: Апрель 8, 2025

As a critical component of rotating machinery, the operating status rolling bearings is not only related to significant economic interests but also has far-reaching impact on social security. Hence, ensuring an effective diagnosis faults in paramount maintaining operational integrity. This paper proposes intelligent bearing fault method that improves classification accuracy using stacked denoising autoencoder (SDAE) and adaptive hierarchical hybrid kernel extreme learning machine (AHHKELM). First, (HKELM) initially constructed, leveraging SDAE's deep network architecture for automatic feature extraction. The functions address limitations single by effectively capturing both linear nonlinear patterns data. Subsequently, (HHKELM) refined through enhanced Aquila Optimizer (AO) algorithm, which iteratively optimizes hyperparameter combination. AO algorithm further incorporating chaos mapping, implementing balanced search strategy, fine-tuning parameter [Formula: see text], collectively improve its ability escape local optima conduct global searches, thus strengthening robustness model during optimization. Experimental results CWRU , MFPT JNU datasets demonstrate autoencoder-adaptive (SDAE-AHHKELM) better accuracy, robustness, generalization than KELM other methods.

Язык: Английский

On the measurement of resonance frequency of nanoclay-reinforced concrete shell structures validated by experimental datasets via artificial intelligence technique and mathematical modeling DOI

Zhonghong Li,

Yang Bing, Suming Chen

и другие.

Measurement, Год журнала: 2025, Номер 248, С. 116961 - 116961

Опубликована: Фев. 7, 2025

Язык: Английский

Процитировано

0
The Segmentation of Tunnel Faces in Underground Mines Based on the Optimized YOLOv5 DOI Open Access

Chundi Ma,

Kechao Li,

Jilong Pan

и другие.

Minerals, Год журнала: 2025, Номер 15(3), С. 255 - 255

Опубликована: Фев. 28, 2025

Tunnel faces in underground mines, as the front line of mining, play an important role both mine safety and mining intelligence. However, engineering quality tunnel is still evaluated based on visual observations by technicians, which cannot guarantee real-time performance. Therefore, there urgent need for a more effective method to detect face engineering. In this study, high-performance accurate segmentation model was developed applying YOLOv5-seg computer vision mine. By optimizing classic Chinese image dataset through Sobel preprocessing improving network structure using SimAM module, good predictive performance achieved segmentation, with values 0.97, 0.89, 0.80, 0.78, respectively, pixel accuracy, Dice coefficient, mask intersection over union (IOU), box IOU test set. And outperforms all YOLOv5 models U-net same task segmentation. Model interpretation visualization further demonstrated positive effect module model, and, finally, results were used evaluate Overall, study’s provide feasible, safe, accurately segmenting mines reliable approach data-driven applications intelligent technology future.

Язык: Английский

Процитировано

0
A computer vision-based real-time monitoring method for swivel bridges spatial rotation DOI
Bei Liu, Ningbo Wang, Can Wang

и другие.

Measurement, Год журнала: 2025, Номер unknown, С. 117236 - 117236

Опубликована: Март 1, 2025

Язык: Английский

Процитировано

0
An adaptive hierarchical hybrid kernel ELM optimized by aquila optimizer algorithm for bearing fault diagnosis DOI Creative Commons
Hao Yan, Liangliang Shang,

Wan Chen

и другие.

Scientific Reports, Год журнала: 2025, Номер 15(1)

Опубликована: Апрель 8, 2025

As a critical component of rotating machinery, the operating status rolling bearings is not only related to significant economic interests but also has far-reaching impact on social security. Hence, ensuring an effective diagnosis faults in paramount maintaining operational integrity. This paper proposes intelligent bearing fault method that improves classification accuracy using stacked denoising autoencoder (SDAE) and adaptive hierarchical hybrid kernel extreme learning machine (AHHKELM). First, (HKELM) initially constructed, leveraging SDAE's deep network architecture for automatic feature extraction. The functions address limitations single by effectively capturing both linear nonlinear patterns data. Subsequently, (HHKELM) refined through enhanced Aquila Optimizer (AO) algorithm, which iteratively optimizes hyperparameter combination. AO algorithm further incorporating chaos mapping, implementing balanced search strategy, fine-tuning parameter [Formula: see text], collectively improve its ability escape local optima conduct global searches, thus strengthening robustness model during optimization. Experimental results CWRU , MFPT JNU datasets demonstrate autoencoder-adaptive (SDAE-AHHKELM) better accuracy, robustness, generalization than KELM other methods.

Язык: Английский

Процитировано

0