Abstract:Accurately extracting faulty sound signals from belt conveyor rollers within the high-noise environment of coal mine operations presents a formidable challenge. To address this issue, this study introduces an innovative fault diagnosis method that merges the variational modal de-composition (VMD) model with the Swin Transformer deep learning network model. First, the study employed the adaptive VMD method to eliminate intense noise from the original signal of the rollers, while also assessing the reconstruction accuracy of the VMD signal across different modal components. Subsequently, we delved into the impact of the parameter structure of the Swin Transformer network model on the fault diagnosis accuracy. Finally, the accuracy of the method was validated using a sound test dataset from the rollers. The results indicated that optimizing the K-value of the VMD method effectively reduced the noise in the reconstructed signal, and the Swin Transformer excelled in extracting both local and global features. Specifically, on the conveyor roller sound dataset, it was shown that, after the VMD reconstruction of the signal so that the highest Pearson correlation coefficient corresponded to a modal component of 3 and adjusting the parameters of the Swin Transformer coding layer, the combination of the VMD+Swin-S model achieved an accuracy of 99.36%, while the VMD+Swin-T model achieved an accuracy of 98.6%. Meanwhile, the accuracy of the VMD+Swin-S model was higher than that of the VMD + CNN model combination, with 95.4% accuracy, and the VMD+ViT model, with 97.68% accuracy. In the example application experiments, compared with other models the VMD+Swin-S model achieved the highest accuracy rate at all three speeds, with 98.67%, 98.32%, and 97.65%, respectively. Overall, this approach demonstrated high accuracy and robustness, rendering it an optimal choice for diagnosing conveyor belt roller faults within environments characterized by strong noise.

Identification Method of Dehydration Stage of Washing Machine based on the Fusion of Sound and Vibration signals

Status Recognition for Fused Deposition Modeling Manufactured Parts Based on Acoustic Emission

Research of Feature Extraction Method Based on Sparse Reconstruction and Multiscale Dispersion Entropy

Vibration Mode Extraction Method Based on the Characteristics of White Noise

In situ monitoring of FDM machine condition via acoustic emission

Unsound Kernel Identification Using Linear Colour Charge‐coupled Device

Appliance Fingerprinting Using Sound from Power Supply

Research on State Recognition Technology of Elevator Traction Machine Based on Modulation Feature Extraction.

Frame vibration states identification for corn harvester based on joint improved empirical mode decomposition - Support vector machine method

A Vibration Feature Extraction Method Based on Time-Domain Dimensional Parameters and Mahalanobis Distance

Combine Assembly Quality Detection Based on Multi-Entropy Data Fusion and Optimized LSSVM

Sound based degradation status recognition for railway point machines based on soft-threshold wavelet denoising, WPD and ReliefF

Sub-Health Identification of Reciprocating Machinery Based on Sound Feature and OOD Detection

Research on Performance Degradation Assessment and Abnormal Health Status Detection of Quayside Crane Lifting Gearbox Based on Information Fusion

Non-intrusive load monitoring algorithm of residential users using limited low-frequency information

Failure Identification Method of Sound Signal of Belt Conveyor Rollers under Strong Noise Environment

A two-stage sound-vibration signal fusion method for weak fault detection in rolling bearing systems

Multi-residual unit fusion and Wasserstein distance-based deep transfer learning for mill load recognition

A VIBRATION RECOGNITION METHOD BASED ON DEEP LEARNING AND SIGNAL PROCESSING

Bearing fault diagnosis of induction machines using VMD-DWT and composite multiscale weighted permutation entropy

Sound Diagnosis Method of Power Transformer Discharge Fault Based on CEEMDAN-SVM