Vestnik MPEI

Development of an Automatic Power Supply Restoration System in Distributed Generation Networks

Mon, 15 Sep 2025 00:00:00 +0000

A tendency toward decreasing the power system inertia caused by a growth in the share of renewable energy sources, taken together with possible unavailability of communication channels, generates the need to develop new approaches to implementing algorithms for emergency control of devices in networks with distributed generation. An algorithm for automatically restoring power supply in such networks is proposed, which relies on local measurements and does not require high-speed communication channels. The algorithm has been validated in an RTDS real-time simulator using a 10.5 kV network model. Two scenarios for checking the algorithm performance were considered: a transition of the network to an island mode resulted from the trip of the main circuit breaker, and a short circuit fault on an outgoing feeder with both successful and unsuccessful automatic reclosing. In the first scenario, the mean time taken to fully restore the power supply was 0.322 s, and in the second scenario this time was equal to 3.637 s in the case of a self-clearing short circuit fault. The simulation results have confirmed the algorithm’s effectiveness in various network operation modes, due to which it holds promise for use in microgrids located in remote areas with minimal operational staff.

The Electric Power System of the Republic of Burundi: the Development History and Current State

Tue, 24 Jun 2025 00:00:00 +0000

The article presents the main stages of the development history and current state of the electric power system of the Republic of Burundi. The main problems relating to the development of the Burundi electric power system and possible ways to solve them are formulated.

Induction-thermal Vacuum Sputtering of Molybdenum Ring Targets Onto Alumina Substrates

Tue, 24 Jun 2025 00:00:00 +0000

The thermal processes associated with induction-thermal vacuum sputtering (ITVS) of ring molybdenum targets are studied. The effect the operating current and the process duration have on the thickness and structure of molybdenum layers deposited by spraying on aluminum oxide plates is determined. An ITVS process numerical model for the “inductor–ring target–sample” system was proposed, using which the temperature field distribution pattern in the sputtered target was determined depending on the inductor current in the range from 1000 to 2500 A and the computer experiment duration up to 600 s. It has been found that by using the above-mentioned modes, molybdenum targets can be heated to a sputtering temperature of 1860–2320°C. In the experimental studies, thermal sputtering of a ring target was carried out using an induction heating unit, special accessories, a two-turn copper inductor at a pressure in the working chamber of 0.6–3.5 Torr, operating frequency f ≈ 45 kHz, and operating current I_O = 1000–1330 A. The process duration was 300–1200 s. It has been shown that the exposure time and the operating current have a directly proportional effect on the coating thickness, with an increase in the operating current having a more intense effect. The thinnest molybdenum layers (h = 2.3±0.3 μm) are produced at the operating current values I_O = 1100 A and exposure time t = 300 s. The maximum thickness of the molybdenum coating is about 22±4 μm at a current of 1300 A and an exposure time of 1200 s. The obtained results provide a wider insight into the thermal sputtering of refractory metals and can be used in development of methods for applying molybdenum layers, including those onto ceramic products for various purposes.

A Review of Photovoltaic System Power Output Forecasting Methods

Tue, 24 Jun 2025 00:00:00 +0000

An integrated temporal and spatial analysis of photovoltaic (PV) system power output forecasting methods plays a key role in ensuring stable power system operation and efficient consumption of renewable energy. The article considers existing studies of PV power forecasting methods from the perspective of different temporal and spatial scales. The forecasting methods have been classified according to the forecasting process, temporal and spatial scales. Based on the temporal and spatial scales of PV power generation, the forecasting methods are put in a systematic order at different temporal scales (ultra-short-term, short-term, and medium- to long-term) and spatial scales (for a single plant and for a region) along with their application scenarios, and technical challenges are discussed. Future development lines are proposed, such as development of simplified models, joint forecasting at different scales, and hybrid methods combining physical modeling and data analysis. The article provides a theoretical framework for selecting and optimizing PV power forecasting methods, aids smart grid management and large-scale application of renewable energy sources, and lays the foundation for further research on state-of-the-art forecasting methods.

Simulation of the KVGM-10 Boiler and the RGMG-10 Burner, when Operating on a Methane-hydrogen Mixture

Mon, 15 Sep 2025 00:00:00 +0000

In recent years, the power engineering community of the leading countries with well-developed economy and energy sectors has focused increased attention on matters concerned with the most environmentally friendly use of existing energy capacities. To solve this problem, it is necessary to modernize the current energy potential of one or another country (and, collectively, their common power system) and develop new principles and technologies for generating, transporting, and consuming energy in its various forms. Hydrogen energy and its derivatives is a topic that is most actively discussed and gaining popularity in different countries, the Russian Federation included. The main objective of hydrogen energy is to use hydrogen in the form of end product as fuel in the energy sector, thereby supporting the climate agenda to reduce the carbon footprint from the impact of the energy sector on the environment. The most promising and discussed vector of national hydrogen power engineering in Russia is the introduction of methane-hydrogen mixtures into the gas transportation sector, namely, the admixing of hydrogen into transported natural gas in a certain concentration. The purpose of the study is to analyze the features pertinent to the operation of Russian gas transportation and gas distribution systems in the case of their possible transition to methane-hydrogen technologies. In carrying out the study, systems analysis and mathematical modeling methods were used. In particular, simulation of the RGMG-10 burner in the ANSYS environment is described. The simulation results have shown that this type of equipment will require certain modification of its design characteristics in the case of its operation on methane-hydrogen mixtures. The furnace unit of a typical boiler installation was simulated in the Aspen Plus environment. The simulation results have shown that during the operation on methane-hydrogen mixtures, a positive tendency toward reducing CO₂ emissions is observed; however, there is also an increase in the concentration of NO_x compounds in the exhaust gases. Research in this field is promising and makes an important contribution to the overall scientific and technical background in the developing segment of Russian hydrogen energy.

Scientific Work of the Department of Low Temperatures: an Overview by the 50th Anniversary. Part 1

Tue, 24 Jun 2025 00:00:00 +0000

An overview of modern scientific works conducted by researchers of the Moscow Power Engineering Institute’s Department of Low Temperatures is presented. Based on the 50-year history and the main directions laid down at the department’s foundation time, advanced theoretical and experimental studies are currently published in leading peer-reviewed publications; they are supported by foundations and ensure the possibility to perform students’ final graduation papers and the defense of dissertations by post-graduate students. Currently, there are five scientific areas of the Department’s researches: technology for producing monodisperse granules from non-ferrous and rare-earth metal alloys; investigation of hydrodynamics and thermal physics of droplet flows; experimental studies of spreading, wetting, and boiling processes in meso- and nanostructures, as well as nanofluids; numerical modeling of evaporation-condensation processes in cryogenic machines and apparatuses; development of methods for calculating heat and mass transfer processes on interfacial surfaces, and their experimental verification on helium-II.

Development and Assessment of the Applicability of a 1D Procedure for Analyzing the Performance of a Radial Turboexpander Operating on Mixed Working Fluids

Tue, 24 Jun 2025 00:00:00 +0000

The article addresses the development and assessment of the applicability of a quasi 1D design methodology for analyzing the performance of a radial turboexpander stage operating on mixed working fluids without considering phase transitions. The applicability of the proposed methodology is assessed by comparing the results obtained using the newly developed approach with the results of CFD calculations carried out in a cyclosymmetric formulation with periodic boundary conditions. Comparison with CFD calculations is adopted in view of complexity of setting up a full-scale experiment, as well as an insufficient scope of materials available in open sources. The developed calculation procedure is based on well-known approaches, but differs from them in offering the possibility of using the equations of state of real gases, considering a mixed working fluid, as well as the transition to the direct setting of isentropic efficiencies of stage elements instead of iterative determination of flow coefficients. This made it possible to simplify the procedure without loss of accuracy. The calculation procedure and its software implementation were tested using the example of constructing the characteristics of a stage for a mixed working fluid (dry air was chosen as an example). A design calculation was carried out, and the geometric and thermal gas dynamic characteristics were determined, which were used as initial conditions for the CFD calculation. Good agreement between the two calculation methods has been obtained; the maximum difference in the characteristics is within 3%, which does not exceed the numerical method uncertainty. Velocity patterns determined using both the approaches (1D calculation and CFD application) have been constructed, the results of which are in good agreement with each other. Based on the CFD calculation results, the pressure and temperature distributions in the radial and meridional sections have been constructed, and the values along the meridional section central line are also given. The consistency of the results with the expanding process physical principles is shown. A non-uniform distribution of temperatures along the height of the guide vane and impeller channels has been revealed. Further development of the procedure is aimed at taking into account possible phase transitions (bulk condensation) in the flow path.

Comparative Analysis of Methods and Approaches for Assessing the Possibility of Steam Explosion and Its Consequences during a Severe Accident at a Nuclear Power Plant with VVER

Mon, 15 Sep 2025 00:00:00 +0000

During a severe accident involving loss of efficient heat removal and subsequent nuclear fuel melting, conditions for the melt to come in contact with water emerge, which can lead to steam explosion (SE). A large-scale SE is a potentially dangerous phenomenon for the nuclear power plant safety, threatening the integrity of the third and fourth protective barriers of the NPP defense-in-depth system, which prevent the spread of fission products. Various methods, methodologies, and approaches have presently been developed to assess the possibility of steam explosion to occur and its consequences during a severe accident at an NPP with VVER. A computational method involving the use of dedicated codes, computational-analytical methodologies, a computational-comparative method, and a phenomenological approach are considered. For qualitatively assessing the steam explosion probability, a phenomenological approach or a computational-comparative method can be used. The phenomenological approach is based on an analysis of studies on steam explosions and the design features of the reactor and the at-reactor spent fuel pool. The computational-comparative method includes two stages: the first stage involves calculations using certified software to obtain initial data; the second stage involves a comparative analysis of the obtained results with the experimental parameter range where a steam explosion is possible. To evaluate the pressure growth and the impact on the NPP structural components during a postulated steam explosion, a computational method with the use of dedicated codes and the computational-analytical methodology can be used. The accuracy of the models embedded in the computational codes is limited by the volume of available experimental data. Within mass, and loads on structural components has been performed based on conservative estimates.

Validating the STEG-IATE Computer Code against the Experimental Data on Hydrodynamics of Two-phase Flows in Vertical Pipes

Александр [Aleksandr] Сергеевич [S.] Никулин [Nikulin] — Tue, 24 Jun 2025 00:00:00 +0000

The aim of the study is to validate the STEG-IATE thermal hydraulic computer code against the experimental data on the hydrodynamics of water-air flows in vertical pipes. A distinctive feature of this code is that it uses the interfacial area transport model (IATE) to determine the drag interaction between the phases. This model is an alternative to the conventional use of a two-phase flow regime map for determining the flow structure and subsequent calculation of the interfacial area. The IATE model advantages are that it predicts the gradual evolution of the interphase area change as a result of the dispersed phase coalescence and fragmentation processes occurring in a two-phase flow. The main difficulty associated with the development and application of this equation is the choice of expressions for describing the dispersed phase fragmentation and coalescence processes, as a result of which a change occurs in the number of particles and, accordingly, the interfacial area. To date, there are quite a lot of models of these phenomena proposed by different authors, but they have not been compared in a systematic manner. In this article, six approaches to describing these processes are considered. The results of validating the STEG-IATE code against the experimental data on two-phase flows in a vertical pipe have shown a noticeable improvement in the predictive ability for the range of low gas and liquid velocities, compared with the previous version of the STEG code with a map of flow regimes. In the range of high gas and liquid velocities, the results obtained using both the codes are close to each other. The testing of approaches to calculating the interfacial area transport has demonstrated that, in general, the best results are given by the model proposed by Smith T.R. It should be noted that the testing results have also shown that there is no universal model. Further improvement of the STEG code will be continued based on the models proposed by Smith T.R., Talley and Hibiki T., and Ishii M.

Development of the Method and Device for Vortex Generation in Fluid Media in Nuclear Reactor Installations

Константин [Konstantin] Николаевич [N.] Проскуряков [Proskuryakov] — Tue, 24 Jun 2025 00:00:00 +0000

A digital model of vortex generation has been developed and verified on the nuclear reactors of the Novovoronezh nuclear power plant.

The Features of Setting up Lighting Installations in Classrooms

Tue, 24 Jun 2025 00:00:00 +0000

The article addresses matters concerned with the operation features of lighting installations taking into account the requirements for natural, artificial and combined lighting of public buildings. Experimental studies on determining the illumination levels on the classroom working surfaces have been carried out; the natural illumination coefficient values have been determined, and the results obtained have been analyzed taking into account the factors affecting the reduction of the operation coefficient. Shortcomings in the operation of lighting installations in classrooms have been identified, which lead to deviations from the requirements established by regulatory documents, and corrective actions to eliminate the shortcomings have been proposed.

Evaluating the Results of Applying Dimensionality Reduction Algorithms for Predicting the Clustering Performance

Tue, 24 Jun 2025 00:00:00 +0000

The possibility of evaluating the effectiveness of different dimensionality reduction methods before the high-dimensional data clustering stage is studied. The main attention is paid to analyzing how well the data passed through different dimensionality reduction algorithms are suitable for subsequent clustering. Different dimensionality reduction algorithms (PCA, Isomap, Locally Linear Embedding, MDS, Spectral Embedding, T-SNE, and UMAP) are applied to four sets of high-dimensional gasoline spectra. The dimensionality reduction results are evaluated using the silhouette coefficient (SC) and the Davies-Bouldin index (DBI). After that, the evaluated data are clustered using different algorithms (DBSCAN, HDB-SCAN, K-Means, Spectral Clustering), and the clustering is estimated using the adjusted Rand index (ARS). Then, the correspondence between the clustering and dimensionality reduction estimates is assessed. Attempts to reliably evaluate most of the dimensionality reduction algorithms using the cluster silhouette coefficient and the Davies-Bouldin index have been met with success. The Chebyshev, Euclidean, Manhattan, and square Euclidean metrics have yielded the most accurate results. Also, the two most effective algorithms T-SNE and UMAP were the only ones that could be evaluated using the DBI assessment. The obtained data can be used to select the optimal dimensionality reduction algorithm prior to perform clustering of multidimensional data, as well as to evaluate the data preprocessing quality. The correlation between the evaluations of dimensionality reduction methods and the efficiency of subsequent clustering has been successfully demonstrated. The T-SNE and UMAP algorithms have shown the best results in evaluation, and the Chebyshev, Euclidean, Manhattan, and square Euclidean metrics have given the most accurate results for these algorithms.

Improvement of the Condensate Temperature Control System for the T-125/150 Steam Turbine of the PGU-450 Combined Cycle Plant

Mon, 15 Sep 2025 00:00:00 +0000

The article considers the problems encountered in heating the main condensate of the steam turbine used as part of the PGU-450 combined cycle plant (CCP) when switching the condenser to operate in the mode with reduced pressure. The condensing unit’s features associated with the need to maintain the main condensate temperature upstream of the gas condensate heater are given. The difficulties associated with heating the condensate by arranging regenerative steam extraction from the steam turbine are noted. A newly proposed method for heating the condensate in a steam-water heater is described. The advantages of the proposed condensate heating method are noted, and the heating system mathematical model is presented along with the algorithm and the results of calculating its efficiency in relation to the conventional condenser operation system during the CCP operation in condensing mode. The lower and upper pressure limits in the condenser, within which the proposed system is recommended for use, are determined according to the energy efficiency criterion. An assessment of the energy and technical and economic efficiency of using the proposed system is given.

Current State and Problems of Developing and Implementing Intelligent Control Systems in Thermal Power Plant Process Control Systems

Mon, 15 Sep 2025 00:00:00 +0000

The article analyzes the current level of development and implementation of intelligent process control systems (PCSs) at thermal power plants (TPPs). The key technological trends, such as the use of digital twins, predictive analytics, and artificial intelligence are considered. The main attention is focused on the systemic, technical, and organizational challenges that hinder the widespread digitalization of power units. A conclusion has been drawn about the need of applying a comprehensive approach combining modern algorithmic solutions with modernizing the infrastructure and personnel training.

Simulating the Operation of an Interconnected System for Controlling the Load and Efficiency of a Stirling Engine-powered Autonomous Installation

Mon, 15 Sep 2025 00:00:00 +0000

Systems for simulating and performing intelligent control of the load and efficiency of an autonomous power installation based on a Stirling engine are presented. The functional diagram, control principles, and algorithms of an intelligent process control system are described, including a sensor unit, an intelligent control unit, an extreme regulator, and actuators. Special attention is given to the automatic operation mode, the system's adaptability, and the use of extreme control and self-diagnostics. The simulation results for the installation operation modes with the extreme controller disabled and enabled are analyzed. It is shown that by using the extreme controller, it becomes possible to decrease fuel consumption, improve efficiency, and provide more stable operation of the installation under the conditions of variable load and various disturbances.

Software Tools for Analyzing Messages of Participants in Public Information Channels

Tue, 24 Jun 2025 00:00:00 +0000

The aim of the study is to develop a method and a chatbot that implements it for public Telegram channels, which will make it possible to delete spam messages from them that are not related to the channel’s main topic or carry advertising information. The main methods for classification and clustering of short text messages to detect spam in them are considered and analyzed. The K-means algorithm, self-organizing maps, spectral clustering, maximum likelihood estimate, topical clustering method, and latent semantic analysis were studied among various clustering methods. Among the classification methods, k-nearest neighbors, Bayes, support vectors, logistic regression, random trees, as well as classification based on artificial neural networks were considered. Based on the analysis results, a solution is proposed that is a combination of the support vector method and a convolutional neural network. Based on the proposed solution, a model of a short text message analyzer was developed, and using the Telegram Bot API interface, a chatbot for the Telegram messenger was developed, which was integrated into real user groups with regular active participants. The testing has shown that the model trained using the support vector method has an accuracy of 77%, and the model trained on the basis of a convolutional neural network has an accuracy of 95%. The study results can be used in administrating the Telegram messenger public information channels. In the chatbot developed on the basis of the proposed classification method, messages from the information channel users are processed in real time, and those that are identified as spam or as not related to the channel topic are deleted. The performance of the developed software has been confirmed by the results of its testing.

An Overview of Hash Function Cryptanalysis Methods

Tue, 24 Jun 2025 00:00:00 +0000

Nowadays, we are witnessing rapid development of digitalization and network technologies, which intensely use cryptographic algorithms like ciphers, hash functions, and digital signatures. To ensure information security, it is important to use strong cryptographic primitives. Their strength is determined by cryptanalysis. The article addresses methods for cryptographic analysis of hash functions. The aim of the study is to develop a method for analyzing cryptographic hash functions based on neural networks. To this end, the following tasks were set forth:

- defining target requirements for cryptographic hash functions for which cryptographic analysis using artificial neural networks (ANN) is possible;

- searching for an algorithm for analyzing hash functions using ANN;

- determining the ANN type for analyzing cryptographic hash functions.

The existing methods for analyzing iterative cryptographic hash functions are classified. A method for analyzing hash functions using neural networks is proposed. The type of artificial neural networks suitable for performing hash function analysis is determined. A method for evaluating the strength of cryptographic hash functions using neural networks is described. The application field of the study results obtained is quite extensive in nature. It includes both the use of ANN to analyze specific hashing functions for resistance to finding a preimage, resistance to searching for collisions of the first and second kind, and the comparison of several hash functions with each other. The use of neural networks in the analysis of hash functions will make it possible to bring the computer analysis of hash functions closer to a cryptanalysis conducted by qualified specialists in the field of information security. Existing methods for assessing the resistance of hash functions to various attacks are based on mathematical principles that may be far from the "real" resistance of a cryptographic primitive. Thus, artificial neural networks provide a different approach to analyzing cryptographic hash functions.

Simulation of a Hidden Data Transmission Channel from an Isolated Workstation in the Ultrasonic Frequency Band

Tue, 24 Jun 2025 00:00:00 +0000

The aim of the article is to study the procedure for qualifying a facility furnished with a workstation based IT features designed to process confidential information (including the identification of hidden information leakage channels). The mechanism through which a hidden acoustic data transmission channel is established is studied in detail, and its simulation is performed using the binary signal modulation method. A computer program simulating malware that generates high-frequency sound waves, and a test bench have been developed based on the study results. Simulation of a hidden ultrasonic data transmission channel has shown its sufficiently high noise immunity and revealed the extent to which the manifestation of unmasking features depends on the transmission rate (manipulation) and on the emitter type.

The Use of Neural Networks to Approximate the Q-function in Reinforcement Learning Methods

Mon, 15 Sep 2025 00:00:00 +0000

The article discusses the possibility of integrating reinforcement learning methods and artificial neural networks. The implementation of Q-function approximation methods in reinforcement machine learning methods is presented. The methods considered are implemented on the basis of a classical neural network and a self-organizing fuzzy neural network. The process through which a self-organizing neural network is automatically tuned is described. The results of testing the approximation methods on the model problems CartPole and MountainCar are presented. The work was carried out within the framework of studying the reinforcement learning methods used in artificial intelligence systems for various purposes, including intelligent real-time decision support systems. The advantages and drawbacks of the Q-function approximation approaches using neural networks in reinforcement machine learning methods are pointed out.