The detection of micro-defects within cemented carbides necessitates a high-frequency, high-sensitivity ultrasonic non-destructive testing transducer (UNDTT), whose performance is highly sensitive to geometric structural parameters. Conventional design approaches rely heavily on empirical trial-and-error, resulting in low efficiency and difficulty in achieving globally optimal solutions. To address this limitation, an intelligent multi-objective optimization method is proposed for transducer structural parameters—namely, radius, matching layer thickness, and backing layer thickness—to simultaneously maximize sensitivity ($V_{pp}$), center frequency ($f_c$), and bandwidth (BW). By investigating the relationship between structural parameters and performance metrics, a dataset was constructed and used to develop a convolutional neural network (CNN) surrogate model that captures their nonlinear mapping. The CNN was integrated with the NSGA-III multi-objective optimization algorithm to iteratively generate a Pareto-optimal solution set, from which the best design was selected using the entropy-weighted Technique for Order Preference by Similarity to an Ideal Solution (TOPSIS). Finite element analysis (FEA) validation confirmed prediction errors below 7.0%. Compared to conventional designs, the proposed approach delivers a 46.1% higher sensitivity and a 7.7% broader bandwidth while maintaining a thinner matching layer. These results confirm the effectiveness and practical advantage of the proposed framework. This data-driven approach offers an efficient alternative for designing a high-performance UNDTT.

To address the challenges of significant size variations, inconspicuous textures, weak edges, and background interference in wire harness insulation surface defect detection, this paper proposes WHD-YOLOv8n, an improved YOLOv8n-based defect detection algorithm that integrates ODConv, CBAM, BiFPN, and Inner-CIoU for adaptive feature extraction, defect-related feature enhancement, multi-scale feature fusion, and bounding-box regression, respectively. WHD-YOLOv8n achieves an mAP@0.5 of 90.15%, representing a 6.21 percentage-point improvement over the YOLOv8n baseline, with Precision, Recall, and F1-score of 92.05%, 89.87%, and 90.95%, respectively. The model has 4.83 M parameters and requires 11.9 GFLOPs, while achieving an inference speed of 135.6 FPS. Ablation experiments and comparisons with representative general-purpose and industrial defect detectors demonstrate the effectiveness of the proposed design and its favorable accuracy–efficiency trade-off. Further experiments under haze and fog degradation show that WHD-YOLOv8n consistently outperforms the baseline under the tested degraded imaging conditions, although severe degradation can still cause missed detections. The proposed method therefore provides an effective solution for wire harness insulation surface defect detection and offers a basis for further research on robust detection under complex industrial imaging conditions.