Search papers, labs, and topics across Lattice.
This paper introduces a semi-supervised bearing fault diagnosis method that leverages a Swin Transformer for feature extraction and an adaptive pseudo-labeling strategy based on prediction entropy to address the scarcity of labeled data. The method employs a two-stage training process: first, training with limited labeled data, and second, incorporating high-confidence pseudo-labels from unlabeled data using adaptive confidence thresholding. Experiments on the Paderborn dataset demonstrate 99.5% accuracy with only 30% labeled data, outperforming conventional methods by 9% and showing robustness to noise and complex faults.
Achieve state-of-the-art bearing fault diagnosis with 30% labeled data by adaptively incorporating unlabeled data using a Swin Transformer and entropy-based pseudo-labeling.
Bearings are critical components of high-speed train bogies, acting as interfaces between rotating wheelsets and the bogie frame. This paper presents a semi-supervised learning approach for bearing fault diagnosis that addresses the challenge of limited labeled data in industrial settings. The method integrates Swin Transformer's feature extraction with an optimized pseudo-labeling strategy, achieving state-of-the-art performance while reducing dependence on labeled samples. A two-stage training process first builds baseline discriminative ability with minimal labeled data, then incorporates high-confidence pseudo-labels from unlabeled data through adaptive confidence thresholding based on prediction entropy. Experiments show 99.5% accuracy on the Paderborn dataset using only 30% labeled data—a 9% improvement over conventional methods. The system excels at identifying complex compound faults and remains robust to noise and data variability. This work also offers theoretical insights into data efficiency in semi-supervised learning, defining optimal thresholds for labeled and unlabeled data use. The approach provides a cost-effective, scalable solution for industrial fault diagnosis while setting new benchmarks for semi-supervised learning in mechanical fault detection.
