Rolling bearings are critical and vulnerable components in mechanical equipment and are prone to various types of damage during operation. Consequently, rolling bearing fault diagnosis is of significant engineering importance. In recent years, deep learning-based approaches have achieved considerable progress in intelligent bearing fault diagnosis. However, existing models still suffer from several limitations, including insufficient feature extraction under noisy conditions, limited diagnostic accuracy, high computational cost, and low operational efficiency. To address these challenges, an intelligent rolling bearing fault diagnosis method based on an ascending-dimensional convolutional neural network (ADCNN) is proposed. Compared with conventional neural networks, the proposed ADCNN features a more compact model size, improved noise robustness, and higher diagnostic accuracy. A large convolutional kernel is introduced in the first layer to enhance noise immunity, while an ascending-dimensional module is employed to reduce the number of network parameters and improve feature extraction capability. In addition, a reduced linear transformation layer (RLTL) is incorporated to further achieve a lightweight architecture. Experimental results on the Case Western Reserve University (CWRU) dataset and a self-designed test dataset demonstrate that the proposed ADCNN achieves superior fault diagnosis performance under different noise environments while maintaining computational efficiency and model compactness.

Multi-source information fusion is regarded as a key approach for improving bearing fault diagnosis. However, due to the heterogeneity of multi-source information, asymmetric information contributions, and imbalanced discriminative features, negative transfer may occur during fusion. To address this issue, this paper proposes a negative-transfer-suppression diagnosis framework based on physical-information guidance and adversarially disentangled representation. First, an adaptive preprocessing mechanism guided by acoustic–vibration cross-correlation and mutual information entropy is constructed to extract intrinsic cross-modal correlations, enabling source-end feature reconstruction and commonality enhancement. Second, an attention-based spatial feature extraction operator and an adversarial common-domain representation model are developed to suppress modality-specific interference and disentangle cross-modal shared features. On this basis, sparse coding is employed to fuse common-domain and modality-specific features. Furthermore, a classification effectiveness evaluation index based on fuzzy clustering is introduced into the loss function to dynamically constrain sparse coding weights, thereby reducing interference features and suppressing negative transfer under strong-noise conditions. Experimental results demonstrate that the proposed method effectively achieves the design objective that “fusion outperforms non-fusion,” exhibiting strong noise robustness and high diagnostic accuracy under complex operating conditions.

In unmanned aerial vehicle inspection of transmission lines, insulators often exhibit arbitrary orientations and elongated shapes and are frequently embedded in complex backgrounds. Horizontal bounding boxes tend to include substantial redundant regions. Meanwhile, existing oriented object detection methods still suffer from angular discontinuities at periodic boundaries, insufficient geometric supervision for rotated bounding boxes, and difficulties in lightweight deployment. To address these issues, this paper proposes a lightweight oriented object detection model, termed R-YOLOv8-PSGH, which integrates dual-frequency phase-shift encoding and Gaussian geometric supervision. Based on a lightweight R-YOLOv8 architecture, a rotated detection head is developed to decouple the predictions of object categories, bounding-box locations, and orientation angles. To improve the periodic continuity of angle representations and strengthen the geometric constraints on rotated bounding boxes, a dual-frequency phase-shift angle encoding strategy and a Gaussian geometric localization loss are designed. Specifically, the complementary relationship between periodic signals with periods of 180°and 90° is exploited to map orientation angles into continuous phase responses, thereby improving the stability of orientation prediction. Moreover, the spatial structure of each rotated bounding box is modeled as a two-dimensional Gaussian distribution, and overlap consistency, center distance, and shape discrepancy are jointly optimized. In this manner, the orientation representation and bounding-box-level geometric supervision are collaboratively enhanced. Experimental results demonstrate that the proposed method improves the detection accuracy and localization stability of rotated objects while maintaining favorable lightweight deployment capability, providing a new solution for lightweight object detection in complex scenarios.