Damian Mazur

The impact of pollution on the performance of photovoltaic (PV) panels and the risk of hot spots is an important issue in the context of optimizing renewable energy systems. Dust, leaves, and bird droppings cause uneven illumination of the panel surface, leading to a decrease in efficiency and local overheating of the cells, which can result in permanent damage. The use of Internet of Things (IoT) technology in PV panel monitoring enables continuous tracking of pollution levels and their impact on system performance. Smart sensors located in various places and on the drones provide real-time data on temperature and air pollution levels of various chemical compounds. The collected information is sent to artificial intelligence-based systems, which analyze patterns and identify potential threats, such as the formation of hot spots or a drop in module performance. Authors method, based on a proprietary structure and selection of deep LSTM network parameters, outperforms other specified machine learning methods in terms of relative prediction accuracy for dust. Proposed algorithm also predicts more accurately than other machine learning methods. Thermal cameras combined with AI algorithms can accurately detect temperature anomalies on the surface of the panels and predict future problems. This allows to optimize cleaning schedule and make maintenance decisions based on actual data rather than periodic inspections.

The increasing energy intensity of the economy has led us to look for ways to reduce this negative trend. One method is non-intrusive load monitoring (NILM). This paper presents the use of artificial intelligence methods for the selection of information features and for the identification of operating electrical devices. A set of potential identification features was obtained from high-frequency measurements covering 12 types of electrical consumers and consisted of 218 features. From these, an identification vector was selected via the mRMR (minimum redundancy maximum relevance) method, which searches for features that are maximally correlated with the class and are as little correlated with each other as possible. Identification was realized by building a hybrid classifier using binary classifiers built from artificial neural networks and decision trees. The Accuracy, Precision, Recall, and F1 metrics were used to assess the quality of identification. The obtained values of the identification quality indicators confirm that it is possible to replace multiclass classification in NILM with binary classification without losing its quality. The use of binary classifiers allows for the identification of new devices without the need to change the classifier configuration.

Unmanned aerial vehicles (drones) are increasingly used in a growing number of applications, both civil and military. Their design is based on low weight, making the presence of shielding a difficult decision between safety and weight. Currently, there are no mathematical models to determine the safety of drones operating near a storm front. Lightning causes an electromagnetic wave of an impulse nature, which may pose a real threat to electronic systems. This work attempts to develop a mathematical model for simulating drone safety in terms of electromagnetic pulses using artificial intelligence-based tools. Actual measurement results collected from four drones were used as training data. They were tested in laboratory conditions using specialized measuring equipment used to test avionics in accordance with international standards such as DO-160. A repeatable surge pulse generator and a data acquisition system allowed us to collect information on how overvoltages propagate inside the drone systems. Systems that directly influence its operation were selected for this purpose, such as the power supply system, engine controllers, GPS, camera and data bus lines. Other works show that most overvoltages are induced in motor coils and antennas. On this basis, a number of formulas and equations were developed to describe the most important elements of the drone, without which its correct operation would not be possible. The results of the analyses and the mathematical model of the drone based on the examined cases are presented in this work as a complement to real experiments.

The increase in photovoltaic systems necessitates addressing hot spot issues. Analysis reveals that shading, dust, and manufacturing defects can lead to hot spot formation. A proposed voltage measurement method, based on the Hot Spot Index (HSI), enables hot spot identification and localization. Experiments confirm its effectiveness in detection. Additionally, nearby surrounding areas are also affected, indicating their potential thermal impact. This proactive approach allows minimizing the adverse effects of hot spots on the performance and lifespan of PV systems.

The paper deals with the electric vehicle (EV) charging challenges amidst electromobility's expansion, highlighting different charging techniques and power control strategies essential for global adoption. It evaluates the effects of charging through simulations in the Technical University of Košice's Electrical Engineering lab, investigating both private and public charging scenarios. The analysis aims to guide future charging infrastructure projects, drawing on practical experiences and consultations with industry experts. Insights gained from the Smart Industry Lab, including access to advanced measuring equipment and software, underscore the study's contributions to understanding EV charging dynamics and its feedback effects on the electrical grid.

The primary objective of this study is to evaluate the accuracy of different forecasting models for monthly wind farm electricity production. This study compares the effectiveness of three forecasting models: Autoregressive Integrated Moving Average (ARIMA), Seasonal ARIMA (SARIMA), and Support Vector Regression (SVR). This study utilizes data from two wind farms located in Poland—‘Gizałki’ and ‘Łęki Dukielskie’—to exclude the possibility of biased results due to specific characteristics of a single farm and to allow for a more comprehensive comparison of the effectiveness of both time series analysis methods. Model parameterization was optimized through a grid search based on the Mean Absolute Percentage Error (MAPE). The performance of the best models was evaluated using Mean Bias Error (MBE), MAPE, Mean Absolute Error (MAE), and R2Score. For the Gizałki farm, the ARIMA model outperformed SARIMA and SVR, while for the Łęki Dukielskie farm, SARIMA proved to be the most accurate, highlighting the importance of optimizing seasonal parameters. The SVR method demonstrated the lowest effectiveness for both datasets. The results indicate that the ARIMA and SARIMA models are effective for forecasting wind farm energy production. However, their performance is influenced by the specificity of the data and seasonal patterns. The study provides an in-depth analysis of the results and offers suggestions for future research, such as extending the data to include multidimensional time series. Our findings have practical implications for enhancing the accuracy of wind farm energy forecasts, which can significantly improve operational efficiency and planning.

The technique of estimating the expected decrease in electricity consumption from the grid and using PV energy for the taken load schedule based on archival data for 5 years is refined. With full self-consumption (SC), the reduction of consumption from the grid can be increased by 9.5%–30.7% for a year according to the rated PV power. Consumption should increase when PV generation exceeds a certain value. A discrete time control of the power of an electric storage boiler (ESB) is proposed based on the deviation of the storage battery (SB) state of charge from a given schedule with a heating concentration during hours of high PV generation. In the considered application, it is possible to increase SC by up to 21%. Reducing the load in the evening allows us to use SB energy to reduce consumption from the grid at night. The possibility of complete photovoltaic SC when the ESB is used with an air conditioner is substantiated. Limitations for air conditioner energy consumption according to PV generation are determined. The system’s 24h model of energy processes is supplemented with a thermal model. The standard use of ESB with water temperature maintenance was also considered for comparison. ESB power control allows you to reduce daily energy consumption from the grid by 1.7–2 times. When combining an adjustable ESB with an air conditioner, it is possible to reduce consumption from the grid by 1.466–1.558 times at minimum and increase consumption from the grid by 2–5% at maximum air conditioner consumption.