The development of smart grids poses great challenges to the scientific and professional community. Increasingly strict requirements from regulators and consumers require appropriate actions from the Distribution System Operator (DSO), infrastructure development, and large investments in the modernization and digitalization of electrical distribution systems. The connection of a large number of electricity sources to the existing distribution grid causes problems that are reflected in unauthorized voltage changes or overloads in the network, as well as compromised power quality. Communication infrastructure, as well as the technologies themselves, are often not satisfactory for the requirements that arise in real networks, and the development of smart grids requires appropriate/advanced information and communication infrastructure. The development of smart grids requires an interdisciplinary approach, experts of different profiles, and clearly defined long-term strategies. This paper provides an overview of existing AI technologies which are proposed for application in power systems, as well as an overview of areas where AI can be implemented to support the operation of power systems in the future (such as maintenance, forecasting, optimization, protection, etc.). In a separate section, a simulation of the production of small PV systems connected to consumer households in weak low-voltage grids (LVG) is presented as an illustrative example. An overview of proposed AI applications in LVGs is provided, along with a discussion of possible improvements and overcoming issues that arise in existing grids with prosumers.

This paper considers the approach for overhead transmission lines’ (OHTL) electric and magnetic field reduction by finding the best phase conductors (PCs) and shield wires (SWs) positions. This approach combines an algorithm based on stochastic modeling for OHTL configuration generation, with the artificial neural networks (ANN) based method for the electric field strength and magnetic flux density determination. This approach enables the generation of an arbitrary number of different OHTL configurations, taking into account specific user-defined limitations. This further, enables to find the OHTL designs that are the best solution for the considered case study regarding electric and magnetic field levels. In this paper, a case study is presented where the considered approach is employed to find the OHTL designs that give the best results regarding the limitation of electric and magnetic field values.

Abstract The methodology for the evaluation of long-term exposure to the overhead line magnetic field is presented, in this paper. The developed methodology is based on the ambient temperature measurements and phase conductors’ height measurements to find a linear regression model to determine phase conductors’ height changes for different ambient temperatures. Based on the overhead transmission line geometry, and datasets about historical overhead line phase current intensity values and ambient temperatures long-term magnetic field exposure can be determined. For magnetic flux density determination, a method based on artificial neural networks is used. The methodology is applied to the case study of overhead line that connect substations Sarajevo 10 and Sarajevo 20. A period of one year is analyzed and magnetic flux density values are determined. The obtained results indicate that during the analyzed period for significant amounts of time magnetic flux density values surpass the recommended values for long-term exposure.

Abstract This paper presents a detailed model of low-frequency oscillations and their damping within the Electric Power System (EPS) of Bosnia and Herzegovina (B&H). The system is modeled using MATLAB software, analysing the steady state and dynamic responses. This research highlights the challenges and impacts of integrating renewable energy sources, such as wind farms, on grid stability and oscillation damping. The paper utilizes eigenvalue analysis to investigate the dynamic characteristics of the system, emphasizing the need for efficient damping strategies to maintain system stability. The methodology includes a comprehensive review of existing literature, the creation of a detailed EPS model of B&H, and the application of eigenvalue and oscillation amplitude analysis to determine damping ratios. The dynamic responses of hydro power plants, HPP Mostar and HPP Jablanica, to transient disturbances are analysed to validate the model and refine damping strategies. The results indicate that the B&H EPS is well-damped, with all eigenvalues possessing negative real parts, and demonstrate the system’s resilience to small disturbances. The results are compared with the technical report on the integration of the wind power plant WPP Podveležje. This comparative analysis shows consistent patterns between the modeled calculations and empirical data, confirming the robustness of the EPS model. This alignment underscores the effectiveness of current damping mechanisms and provides a foundational strategy for enhancing system stability with increasing renewable energy penetration. The findings highlight the importance of developing advanced control strategies to sustain system stability as the integration of variable renewable energy sources continues to grow.

Abstract This paper investigates the potential application of neural networks for predicting electricity production in hybrid systems combining photovoltaic (PV) panels and wind turbines. The research focuses on identifying key factors affecting the efficiency and reliability of these systems, including weather variability, PV panel temperature control, solar irradiation, and panel contamination by dust and other pollutants. Artificial neural network (ANN) models are used to predict power output, incorporating robust data filtering and parameter optimization techniques. Through case studies from Germany, the significant role of stochastic weather patterns on energy production is demonstrated, highlighting the need for accurate modeling and strategic management. The findings emphasize that accurate modeling and prediction are crucial for optimizing the operation and reliability of hybrid systems, facilitating a reduced dependency on fossil fuels and promoting sustainable power accessibility in remote areas. By applying a Feed Forward Back Propagation Network (FFBPN), this research demonstrates improved prediction accuracy of power outputs, which is crucial for effective integration and management of renewable sources in the power grid. The study supports ongoing refinement of predictive models and system integration strategies to fully harness the potential of hybrid renewable energy systems.

The paper presents an algorithm for determining the optimal connection location and power of a photovoltaic plant in a distribution network. The proposed algorithm is based on the use of the fuzzy logic and power flow calculation method. The fuzzy logic is used for the selection of candidate buses for the photovoltaic plant connection, while load flow analysis is used for the verification of voltage conditions and power losses in the distribution network. For each of the candidate buses photovoltaic plant of a certain power range was considered. The practical application of the considered algorithm was demonstrated on a part of Sarajevo's 10 kV distribution network.

Due to the significant growth in the number of devices, the range of services it provides, and strict air conditioning requirements, the telecommunications infrastructure is becoming an increasingly important electricity consumer. The efficiency of the power supply system and the power quality are significant challenges in the design and maintenance of telecommunications infrastructure elements. In such systems, power electronic converters play an indispensable role. This paper discusses the results of power quality measurements for supply systems of telecommunications devices. The power supply systems of telecommunications devices with different power converters were analyzed. Also, the power supply of a mobile telephony base station at a remote location was considered, with special reference to the reaction of battery storage in the event of a power outage. Obtained results demonstrate that it is necessary to treat such consumers with special care and take measures to limit their emission of current harmonics.

This paper presents the use of the Hilbert-Huang Transform (HHT) to identify low-frequency electromechanical oscillatory modes, their characteristics, and damping. As these oscillations can have varying features, locations, and impacts on power systems, identifying and monitoring them is crucial for the monitoring, protection, and control of modern power systems. The Hilbert-Huang transform (HHT) is a technique used to analyze nonlinear and non-stationary time series data. It involves breaking down the data into components using Empirical Mode Decomposition (EMD), which generates components with varying amplitudes and frequencies. The EMD process includes an inner loop called sifting, which produces an Intrinsic Mode Function (IMF) until the signal reaches a mean value of zero or a maximum number of iterations. The obtained IMF is a characteristic function of a fundamental oscillation that is symmetrical around the abscissa. The dominant oscillatory mode's frequency can be determined by applying the Hilbert transformation to the first IMF, and the damping ratio and damping can be calculated by fitting a least square line to the logarithmic instantaneous amplitude of the first IMF. To demonstrate the efficacy of the methodology, three case studies are examined. The first case involves generating a synthetic signal to simulate a load angle change with a defined frequency and damping. In the second case, a small disturbance in mechanical power change in the Single Machine System is simulated. The third case simulates a three-phase short circuit on the transmission line using the IEEE 39 bus test system. The results are compared to modal analysis conducted in DigSilent PowerFactory software. The application of HHT yielded satisfactory and promising results in identifying the dominant mode's oscillation frequency and damping.

Abstract Underfrequency load shedding is a common technique for maintaining the stability of the power system by removing the overload in a certain part of the system after a disturbance. The purpose of underfrequency load shedding is to balance output and load when a particular event causes a significant frequency drop in the power system. In conventional underfrequency load shedding schemes, the frequency thresholds of frequency relays are constant, this way it is difficult and sometimes impossible to control the frequency in various disturbances in the system. In this paper, an adaptive underfrequency load shedding (AUFLS) algorithm that is independent of communication between relays is presented. The relays are tuned to reduce loads taking into account local parameters such as voltage and frequency to prevent the occurrence of a cascade failure that can ultimately lead to the breakdown of the entire power system. In this paper, the rate of change of frequency (ROCOF) is obtained by applying the Hilbert-Huang transformation. Numerical simulations conducted on the New England 39 bus test system in the DIgSILENT PowerFactory and MATLAB software packages confirm the effectiveness of the proposed approach.