During the maintenance and management of solar photovoltaic (PV) panels, how to efficiently solve the maintenance difficulties becomes a key challenge that restricts their performance and service life. Aiming at the multi-defect-recognition challenge in PV-panel image analysis, this study innovatively proposes a new algorithm for the defect detection of PV panels incorporating YOLOv7-GX technology. The algorithm first constructs an innovative GhostSlimFPN network architecture by introducing GSConv and depth-wise separable convolution technologies, optimizing the traditional neck network structure. Then, a customized 1 × 1 convolutional module incorporating the GAM (Global Attention Mechanism) attention mechanism is designed in this paper to improve the ELAN structure, aiming to enhance the network’s perception and representation capabilities while controlling the network complexity. In addition, the XIOU loss function is introduced in the study to replace the traditional CIOU loss function, which effectively improves the robustness and convergence efficiency of the model. In the training stage, the sample imbalance problem is effectively solved by implementing differentiated weight allocations for different images and categories, which promotes the balance of the training process. The experimental data show that the optimized model achieves 94.8% in the highest mAP value, which is 6.4% higher than the original YOLOv7 network, significantly better than other existing models, and provides solid theoretical and technical support for further research and application in the field of PV-panel defect detection.

To address the limitations of the traditional long-vector model, which cannot adequately preserve the inherent orthogonal structure among channels in the signal processing of a single vector hydrophone, three minimum variance distortionless response (MVDR) direction-of-arrival estimation algorithms based on the quaternion model are proposed for a single MEMS vector hydrophone. Compared with the traditional long-vector model, the quaternion can retain the structural relationship between channels in a more compact data representation form. First, based on the output characteristics of the sound pressure and dual-channel vibration velocity of the two-dimensional MEMS vector hydrophone, three quaternion signal models—namely Q-VV, Q-PVV, and Q-PV—are constructed. Subsequently, the corresponding MVDR spatial spectra are derived in the quaternion domain. To enhance the direction-of-arrival resolution performance under low-signal-to-noise-ratio (SNR) conditions, the algorithms are then improved by incorporating subspace decomposition theory. The simulation results show that the three proposed quaternion-based MVDR algorithms can all realize effective direction-of-arrival estimation for a single target. When the signal-to-noise ratio is greater than −4 dB, the direction estimation error is less than 1°. The results of dynamic bearing trajectory tests and fixed-point direction-of-arrival estimation experiments indicate that the proposed method can realize the continuous and stable tracking of the target azimuth. Among the involved algorithms, Q-PV-MVDR can effectively suppress false targets, with an average relative angle orientation error of 0.5°.

To improve the stability of ohmic contacts for SiC sensor chips operating at 600 °C, SiC/Ti/TaSi₂/Pt ohmic contact stacks were fabricated on n-type 4H-SiC in this work. The effects of the Ti/TaSi₂/Pt thickness ratio and annealing conditions on electrode performance were systematically investigated, and the resulting electrodes were characterized via SEM, XRD and AES. The results show that a lower specific contact resistivity (SCR) (7.16 × 10⁻⁵ Ω·cm²) is obtained after the Ti/TaSi₂/Pt ratio of 1:3:2 (50/150/100 nm) is rapidly annealed at 1050 °C. The SCR remains 3.05 × 10⁻⁴ Ω·cm² after 20 h of atmospheric aging at 600 °C, and the electrode surface remains flat enough for the subsequent processes. Taken together, the XRD and AES results suggest that the formation of well-crystallized TiC/TaC and Si-enriched titanium silicides lowers the specific contact resistance, and that the thermally stable phases, such as TaC, hinder oxygen ingress, which is consistent with the observed high-temperature stability. In contrast, oxygen penetrates into the contact interface and forms TiO₂, an insulating oxide that increases the contact resistance and thereby degrades the contact performance. The results provide a basis for the design of ohmic contacts for high-temperature silicon carbide sensors.