Accurate solar power forecasting is a key technology for efficient operation of photovoltaic (PV) power plants and safe grid dispatch. Under the “dual carbon” goals and the increasing share of renewable energy connected to the grid, ultra-short-term power forecasting is important for improving dispatch decisions and supporting system operation. To address the ultra-short-term forecasting task at two PV sites, this study develops an end-to-end framework that integrates machine learning, Bayesian optimisation, and SHAP-based interpretability. First, correlation analysis was performed on the datasets from the two sites to provide a foundation for subsequent model development. Next, seven forecasting models, including CatBoost, NGBoost, Random Forest (RF), AdaBoost, ARIMA, CNN-LSTM, and LSTM, were developed and uniformly optimised using Bayesian optimisation. Under a unified framework of data partitioning, optimisation budget, and evaluation metrics, the predictive performance of all models at the two sites was systematically assessed. The results show that the optimal model varies across sites: at Site 1, LSTM delivered the best performance, with test-set R², MSE, RMSE, and MAE values of 0.972, 17.610, 4.196, and 2.267, respectively; at Site 2, CatBoost achieved the best results, with corresponding values of 0.994, 0.385, 0.621, and 0.249, respectively. These findings highlight pronounced site-specific differences in model performance, indicating that different modeling approaches exhibit distinct adaptability under varying data characteristics and operational conditions. Further error analysis and SHAP interpretation indicate that solar irradiation and key meteorological variables are the main drivers of power output, and their effects are nonlinear, confirming the model’s ability to capture complex nonlinear relationships in PV power forecasting. Finally, a graphical user interface (GUI) tool was developed to support site selection, real-time forecasting, and parameter input, providing a practical and convenient solution for PV plant operation and grid dispatch.

Low-altitude UAV RGB-T small-object detection is challenged by unequal modality reliability and the progressive attenuation of small-object evidence. To address these issues, this paper proposes Modality-Dominant Controlled Fine-Tuning with Hierarchical Small-Object Modeling (MDCFT-HSM), an ordered feature-flow framework for RGB-T detection. Based on dataset-level single-modality performance, MDCFT selects an initial protected branch and introduces auxiliary-modality information through zero-initialized residual mappings at the stem, P3, P4, and P5 stages, thereby limiting interference from degraded auxiliary features. Within this controlled feature flow, the High–Low Frequency Detail Enhancement module (HLFDE) preserves shallow boundary, texture, and local thermal-response cues; the Selective Boundary-Guided Aggregation module (SBGA) strengthens cross-level detail propagation in the neck; and the Multi-scale Guided Feature Recalibration module (MGFR) recalibrates the P3, P4, and P5 features before the Detect head. On RGBTDronePerson, MDCFT-HSM achieves 49.53% mAP@0.5 and 18.97% mAP@0.5:0.95 with 7.38 M parameters, 44.2 GFLOPs, and an inference speed of 60.0 FPS. On DroneVehicle, it achieves 83.60% mAP@0.5 and 63.28% mAP@0.5:0.95. Controlled robustness tests show limited tolerance to auxiliary RGB absence and mild cross-modal spatial misalignment, whereas severe IR degradation, IR absence, and larger spatial offsets cause substantial performance loss. These results demonstrate a competitive accuracy–complexity trade-off under a fixed dataset-level protected-branch configuration. The method does not provide online sample-level reliability adaptation or geometric registration.

Accurate wind power forecasting is essential for efficient wind farm operation and reliable grid dispatch. This study proposes a site-adaptive forecasting framework that integrates machine learning, Tree-structured Parzen Estimator (TPE)-based Bayesian hyperparameter optimization, and SHapley Additive exPlanations (SHAP) for interpretability. Using real-world meteorological and power generation data from two wind farms, we first perform joint-distribution feature analysis to characterize statistical relationships between key inputs and power output, supporting model development and interpretation. TPE optimization is then applied to six benchmark models (CatBoost, Extra Trees, GBM, LightGBM, TabNet, and XGBoost). The optimized Extra Trees model achieves the best performance at Site 1 (R² = 0.965, RMSE = 3.872 kW, MAE = 2.333 kW), whereas the optimized XGBoost model performs best at Site 2 (R² = 0.921, RMSE = 3.049 kW, MAE = 1.382 kW), demonstrating the effectiveness of TPE tuning and the strong predictive capability of tree-ensemble learners. SHAP analysis further reveals heterogeneous drivers across sites: Site 1 benefits from synergistic wind-speed contributions across multiple heights, while Site 2 is primarily governed by hub-height wind speed. Overall, the proposed framework achieves both high accuracy and robust interpretability for multi-site wind power forecasting.

In low signal-to-noise ratio (SNR) environments, shortwave channels are subject to impairments such as multipath effects, frequency-selective fading, and time-selective fading. These significantly increase the bit error rate (BER), adversely affecting communication quality and making it difficult to meet communication requirements. Addressing the limitations of traditional techniques like the Least Squares (LS) and Minimum Mean Square Error (MMSE) methods, this paper proposes a deep learning-based demodulation algorithm for shortwave channels. The algorithm employs a multi-channel approach, feeding the real and imaginary parts of the signal as separate input channels to enhance demodulation accuracy. It integrates a convolutional neural network (CNN) module for local feature extraction, aiming to more effectively capture the signal’s local characteristics. Simulation results demonstrate that within the SNR range of -10 dB to 10 dB, the deep learning approach significantly outperforms traditional methods in demodulation performance. Compared to benchmark methods including LS channel estimation, MMSE channel estimation, and other networks such as CNN, GRU, and CNN-RNN, the proposed Transformer-based network demodulation method consistently achieves SNR gains of 1–5 dB. The simulations prove that the proposed Transformer-based demodulation receiver exhibits superior performance under low-SNR, high-frequency channel conditions compared to both traditional methods and existing deep learning models. These findings offer valuable insights for enhancing the reliability and stability of shortwave communication systems.

For power-grid applications such as transmission corridor inspection, substation asset inspection, and post-disaster emergency repair, reliable UAV self-localization under GNSS-degraded or GNSS-denied conditions is critical to ensuring operational safety and accurate defect geotagging. Due to substantial discrepancies in viewpoint, scale, and geometric structure between oblique UAV images and nadir satellite images, conventional RGB-based cross-view retrieval methods often suffer from unstable alignment and insufficient geometric modeling, particularly in scenarios with repetitive textures and partial overlap. To address these challenges, we propose a cross-view visual geo-localization model that integrates RGBD multimodal inputs with multi-scale attention enhancement. Specifically, MiDaS is used to estimate relative depth from UAV imagery, which is concatenated with RGB to form a four-channel input, while satellite images are padded with an additional zero channel to maintain dimensional consistency. A shared-weight ViTAdapter is adopted to learn joint semantic–geometric representations, and a lightweight Efficient Multi-scale Attention (EMA) module is adopted on spatial feature maps to strengthen multi-scale spatial consistency. In addition, an IoU-weighted InfoNCE loss is employed to accommodate partial matching during training, thereby improving the robustness of feature alignment. Experiments on the GTA-UAV dataset under the cross-area protocol show stable performance across both retrieval and localization metrics. Specifically, Recall@1, Recall@5, and Recall@10 reach 18.12%, 38.83%, and 49.47%, respectively; AP is 28.01 and SDM@3 is 0.53; meanwhile, the top-1 geodesic distance error Dis@1 is 1052.73 m. These results indicate that explicit geometric priors combined with multi-scale spatial enhancement can effectively improve cross-view feature alignment, leading to enhanced robustness and accuracy for localization in challenging power inspection scenarios.

Lightweight detectors for steel surface defects still struggle to balance robust feature representation with deployment efficiency when defects exhibit weak textures, fine details, and large variation in scale and shape. To narrow this gap, this study develops an incremental lightweight enhancement of YOLO11n rather than proposing a brand-new detection framework. The method combines two coordinated changes. First, an improved lightweight channel-aware fusion module with a dilated branch, termed LCAF-D, is inserted into the neck to strengthen context-aware multi-scale aggregation at low additional cost. Second, a hierarchical Mish activation strategy is introduced only in selected deep backbone layers and neck downsampling layers so that nonlinear modeling is strengthened without disturbing shallow low-level feature extraction. Experiments on NEU-DET indicate that the proposed design provides competitive lightweight detection performance. The benefit becomes more evident as the input resolution increases. Under the 640 × 640 setting, YOLO11n + LCAF-D + Mish reaches 77.1% mAP@0.5 and 44.3% mAP@0.5:0.95, exceeding the YOLO11n baseline. Five-seed repeated experiments further show slightly better mean accuracy with stable variation, and runtime benchmarking shows that the method keeps lightweight characteristics with only modest increases in parameters, GFLOPs, latency, and GPU memory. Additional evaluation on GC10-DET under the same 640 × 640 protocol also gives an overall improvement over the baseline, although the gains remain category-dependent rather than universal. Overall, the method is best understood as a competitive lightweight engineering refinement for steel surface defect detection, particularly when the input resolution is sufficient to preserve subtle defect details.