The aim of the presented investigation is to explore the time gap between an anomaly appearance in continuously measured parameters of the device and a failure, related to the end of the remaining resource of the device-critical component. In this investigation, we propose a recurrent neural network to model the time series of the parameters of the healthy device to detect anomalies by comparing the predicted values with the ones actually measured. An experimental investigation was performed on SCADA estimates received from different wind turbines with failures. A recurrent neural network was used to predict the temperature of the gearbox. The comparison of the predicted temperature values and the actual measured ones showed that anomalies in the gearbox temperature could be detected up to 37 days before the failure of the device-critical component. The performed investigation compared different models that can be used for temperature time-series modeling and the influence of selected input features on the performance of temperature anomaly detection.

The topology of the static synchronous compensator of reactive power for a low-voltage three-phase utility grid capable of asymmetric reactive power compensation in grid phases has been proposed and analysed. It is implemented using separate, independent cascaded H-bridge multilevel inverters for each phase. Every inverter includes two H-bridge cascades. The first cascade operating at grid frequency is implemented using thyristors, and the second one—operating at high frequency is based on the high-speed MOSFET transistors. The investigation shows that the proposed compensator is able to compensate the reactive power in a low-voltage three-phase grid when phases are loaded by highly asymmetrical reactive loads and provides up to three times lower power losses in the compensator as compared with the situation when the compensator is based on the conventional three-level inverters implemented using IGBT transistors.

This study aims to comprehensively analyze five weather forecasting models obtained from the Open-Meteo historical data repository, with a specific emphasis on evaluating their impact in predicting wind power generation. Given the increasing focus on renewable energy, namely, wind power, accurate weather forecasting plays a crucial role in optimizing energy generation and ensuring the stability of the power system. The analysis conducted in this study incorporates a range of models, namely, ICOsahedral Nonhydrostatic (ICON), the Global Environmental Multiscale Model (GEM Global), Meteo France, the Global Forecast System (GSF Global), and the Best Match technique. The Best Match approach is a distinctive solution available from the weather forecast provider that combines the data from all available models to generate the most precise forecast for a particular area. The performance of these models was evaluated using various important metrics, including the mean squared error, the root mean squared error, the mean absolute error, the mean absolute percentage error, the coefficient of determination, and the normalized mean absolute error. The weather forecast model output was used as an essential input for the power generation prediction models during the evaluation process. This method was confirmed by comparing the predictions of these models with actual data on wind power generation. The ICON model, for example, outscored others with a root mean squared error of 1.7565, which is a tiny but essential improvement over Best Match, which had a root mean squared error of 1.7604. GEM Global and Gsf Global showed more dramatic changes, with root mean squared errors (RMSEs) of 2.0086 and 2.0242, respectively, indicating a loss in prediction accuracy of around 24% to 31% compared to ICON. Our findings reveal significant disparities in the precision of the various models used, and certain models exhibited significantly higher predictive precision.

In this study, we investigate how computationally simplified activation functions affect predictive performance, inference latency, and energy usage in long short-term memory-based temperature prediction for wind turbine generator bearings. We tested three different types of long short-term memory (LSTM) cells, along with bidirectional LSTM (biLSTM) networks, to determine their effectiveness in modeling dynamic changes in gearbox bearing temperatures. We compared several activation-function variants, focusing on variants that are either computationally simple or known to give good performance in deep recurrent networks. The results show that the best-performing architectures achieved root mean squared errors (RMSEs) between $0.0798$ and $0.0822$, corresponding to coefficients of determination in the range of $R^2=0.84$–$0.85$. When applied across five turbines, the best-performing architectures (peephole and bidirectional) achieved root mean squared errors of $0.0898$, $0.0882$, and $0.042$, respectively. The best activation function-enhanced variant (the peephole) improved accuracy by approximately $3\%$ while maintaining low model complexity. These findings provide a practical and efficient solution for embedded predictive maintenance systems, providing high accuracy without incurring the computational cost of deeper or bidirectional architectures.

Deepfake detection from images and videos has evolved from artifact-specific convolutional baselines toward more generalizable, cross-dataset, and foundation-model-based approaches. The current work focuses on the efficiency and informativeness of frame selection itself, while keeping the downstream detectors fixed. The study compares twelve frame-selection heuristics ranging from simple baselines to landmark-aware strategies. Four pre-trained detectors were included in the present quantitative comparison: Self-Blended Images (SBI), Frequency-Enhanced Self-Lendered Images (FSBI), Generative Convolutional Vision Transformer (GenConViT), and GenD. The results show that GenD achieved the strongest average detector-level performance, with a mean frame-mean AUC of 0.9464. The best single validated configuration is GenD, yielding an AUC value of 0.9607 and a balanced accuracy of 0.9133. FSBI and SBI reached mean AUC values of 0.8953 and 0.8935, respectively, while GenD was the best general candidate. For SBI, the best validation configuration is Landmark cluster with 32 selected frames. GenD achieves the best AUC at the level of selection strategy. The present work demonstrates that inference-time frame selection is an important component of video-only deepfakes under constrained inference budgets.

This study evaluates frame selection during inference as an independent factor in video deepfake detection while keeping the downstream detectors fixed. We compare twelve frame selection strategies, ranging from simple temporal and quality baselines to landmark aware policies, using four validated pretrained detectors: Self-Blended Images (SBIs), Frequency-Enhanced Self-Blended Images (FSBIs), Generative Convolutional Vision Transformer (GenConViT), and GenD. The primary experiment is a complete factorial benchmark with 300 videos and five frame budgets (2, 4, 8, 16, and 32 selected frames), which provides the reference results at 32 frames. To address sample size limitations, an additional validation experiment uses a deduplicated split of 1180 Celeb-DF++ and FaceForensics++ videos, with complete results for 2, 4, and 8 selected frames and a reported subset for 16 selected frames. In the complete 300-video benchmark, 32 frames achieved the strongest average AUC, while 8 and 16 frames recovered most of the attainable performance with lower runtime. The best single validated configuration was GenD with Shot-aware sampling at 32 frames, yielding an AUC of 0.9607 and a balanced accuracy of 0.9133. The study therefore does not claim that smaller budgets universally outperform 32 frames; instead, it quantifies the tradeoff between accuracy and runtime and shows that frame selection remains a meaningful design variable under constrained inference budgets.