Damage Detection Method for High-Strength Aramid Conveyor Belts

Abstract

Mining conveyor belts play a crucial role in the mining transportation system, and the detection of damage caused to them can affect productivity and safety. Since the traditional methods of detection perform poorly in efficiency and accuracy, a new method of damage detection is proposed based on the FCSA - YOLOv10 model in this paper for high-strength aramid conveyor belts. This method is used to capture the image of aramid conveyor belts by irradiating the conveyor belt with X-rays, and the FCSA-YOLOv10 model is applied to damage detection. By introducing the Faster structure, the C2f structure is improved. Also, the Convolutional Block Attention Module (CBAM) is used to replace the Multi-head Self-Attention (MHSA) in the PSA structure, while the SimSPPFCSPC module is used to replace the original SPPF structure. Meanwhile, the Adaptive Threshold Focal Loss Function (ATFL) is applied to address the imbalance between the background and the target in images of the aramid conveyor belt. Experimental results show that the FCSA-YOLOv10 model significantly improves the accuracy of detection, recall rate, and precision on the dataset of mining aramid conveyor belt. Specifically, they reach 98.5%, 95.1%, and 96.7%, respectively. To sum up, this research contributes an effective solution to the efficient detection of damage caused to high-strength aramid conveyor belts.

Keywords

• Aramid conveyor belt
• non-destructive testing
• X-ray
• FCSA-YOLOv10
• SimSPPFCSPC module