Background . Improving the metrological characteristics of measuring instruments for parametric sensors parameters in the development of new methods for measuring their parameters, represented as a two (three) or more element equivalent circuit in the form of a two-terminal electrical circuit with a simple circuit implementation. The aim of the work is to develop a method for measuring sensor parameters regardless of the input parameters with the ability to control the measurement error. Materials and methods . Metrological characteristics of the method were studied using analytical modeling in MatLab/Simulink, allowing for a reduction in time and elimination of errors at the initial stages of measuring instruments design. Results . As a result of analytical calculations, relationships were derived for measuring sensor parameters and an assessment of the measurement error of the proposed method was carried out.
Background. This article presents methods for assessing the condition of one of the most relevant types of robotic systems - unmanned aerial vehicles (UAVs). The objective of the study is to fine-tune the data methods and conduct a brief comparative analysis. Materials and methods. The main methods of the study were literature review and computer modeling. Results. The following methods for assessing the condition of RS (as applied to UAVs) are considered: the complementary filter, the Kalman filter, the Madgwick filter, and data processing using artificial intelligence. Conclusion. The CF used for manually controlled UAVs is the fastest, the Kalman filter is the most advanced, and the Madgwick filter is a compromise solution guaranteed for use in autonomous UAVs. AI-based solutions consume too many computing resources with such low efficiency.
Background. This paper examines modern approaches to data structuring and information enrichment in the context of digitalization. Approaches to solving the problem of automated data extraction are proposed using the example of tool manufacturers' catalogs in order to minimize manual labor, reduce time costs and apply them in the «TKMP.Istok» marketplace. Materials and methods. Methods for detecting objects using the YOLOv10 model and optical character recognition using the EasyOCR and PaddleOCR models are considered. Results and conclusions. As a result of the experimental application, the processing time of documents has been reduced by more than 90 %. The methods used make it possible to obtain high accuracy of information extraction, which opens up new prospects for the use of artificial intelligence in business processes as a key component of digital transformation.
Background . The aim of this work is to improve automated diagnostic processes and enhance the accuracy of early screening for Alzheimer's disease using electroencephalographic signals. Materials and methods . The research method is based on the implementation of algorithms for electroencephalographic signals conversion and a preliminary assessment of the predictors using the Support Vector Machine method. The set of source signals used in this study was taken from the open-access OpenNeuro database. Results . The results of the work include the development of software components for an electroencephalographic signal conversion system, designed to extract a set of classification features (predictors). These predictors are used to determine a subject's belonging to a specific class via machine learning methods. Furthermore, recommendations are formulated for the application of the proposed approach in the design and software implementation of diagnostic systems for Alzheimer's disease, as part of larger software and hardware complexes. Conclusions . An algorithm for converting electroencephalographic signals and extracting a set of features that enable training a machine learning model for classification and diagnosis of Alzheimer's disease has been developed and implemented.
Background. The maximum production of functional simple two-layer and complex multi-layer printed circuit boards is based on technological processes and the availability of production capacities. The existing assessment of the maximum production based on production capacities has low relevance, which makes it difficult for manufacturing enterprises to use it. It is necessary to develop a methodology for assessing production capacities for PCB manufacturing enterprises. Materials and methods . A methodology for assessing production capacity using mathematical tools is presented. Results . The methodology for assessing production capacity has been developed in terms of such organizational and technical system flows as the production of double-layer printed circuit boards and the production of multi-layer printed circuit boards. Conclusions. The developed methodology for assessing production capacity has been tested at a printed circuit board manufacturing enterprise, and further research tasks have been identified.
Background. The relevance of this research is related to the increasing requirements for the reliability and stability of magnetic field sensors, which requires improving the electrical parameters of magnetoresistive elements. Materials and methods. An engineering calculation algorithm has been developed and described in detail, including the selection of a material (Fe20Ni80 alloy), the determination of geometric parameters taking into account the technological limitations of photolithography, the optimal number of links, and the assessment of manufacturing errors. A distinctive feature of the methodology is the refined account of the influence of the shape of the bends on the total resistance through the introduction of an effective shape factor for trapezoidal sections ( K ф ≈ 0.18). Results. A new configuration of magnetoresistive element topology with 45° trapezoidal bends is presented, providing a compromise between manufacturability and reduced current density gradient. Conclusions: Using a magnetoresistor made of Fe20Ni80 alloy ( R = 2.3 kOhm, P = 50 mW), the method's effectiveness and the possibility of implementing the element in the temperature range of -40…+65 °C have been proven.
Background. The urgent tasks facing modern industry and the energy complex are to increase the energy efficiency of existing systems and switch to renewable and environmentally friendly energy sources. The technology of electric energy production using mini-turbines, which use the energy of the flow of freely moving wastewater from wastewater disposal systems, is aimed at solving these problems. The main objectives of the work are to evaluate the operability and efficiency of the considered technology for the production of electric energy using mini-turbines using simulation modeling; to develop an effective structure for controlling the power supply process based on an information measuring and control system. Materials and methods. Mathematical modeling and computational fluid dynamics methods were used to simulate the operation of a mini-turbine. When developing a power supply process control system, methods of analysis and synthesis of systems in multiple states of operation, the theory of optimal control, principles and methods of building information measurement and control systems were used. Results. A three-dimensional model was constructed and a simulation of the functioning of a mini-turbine in the pipeline of the wastewater disposal system was carried out. The scheme of connection and interaction of the main elements of power supply and control systems is proposed. As a key element of the power supply process control system, it is proposed to use an information measuring and control system, which is a software and hardware complex that provides solutions for monitoring technological parameters and optimizing the operating modes of the power supply system equipment. Conclusions. The results obtained confirmed the operability and efficiency of the considered technology for the production of electric energy using mini-turbines. The use of mini-turbines as an additional source of electric energy generation makes it possible to increase energy efficiency by reducing the consumption of electric energy from the external network.
Background. The article addresses the pressing issue of automating the control of design documentation (EDS - Unified System of Design Documentation). Existing approaches, including rule-based systems and the isolated use of machine learning models, demonstrate limited effectiveness in identifying semantic contradictions between drawings and specifications. Materials and methods. As a solution, a hybrid architecture is proposed that implements cross-modal analysis of textual and graphical information using modified RuBERT and Cascade R-CNN models. The scientific novelty lies in the development of an algorithm for interfacing heterogeneous feature spaces, which ensures verification of the semantic consistency of documents. Results . Comparative testing of four approaches was conducted during the experiment on a sample of 100 specification-assembly drawing pairs. The hybrid model demonstrated superiority over all analogues, achieving an accuracy (Accuracy) of 95.0 %, detection precision (Precision) of 94.1 %, recall (Recall) of 91.4 %, and an F1-measure of 92.7 %. In contrast to the rule-based system (F1-Score = 0.00), isolated CNN (F1-Score = 73.5 %), and isolated RuBERT (F1-Score = 86.9 %), the proposed solution provides comprehensive control by detecting cross-modal inconsistencies. Conclusions . The results confirm the feasibility of implementing the developed methodology into PLM system contours to reduce the labor intensity of standard control and minimize errors that go into production.
Background . The purpose of the work is to study models of the flow tube of a Coriolis flow meter with varying degrees of abrasive wear and evaluate its effect on the accuracy of determining mass flow parameters. Materials and methods. Simulation models make it possible to assess the influence of various parameters on the operation of a Coriolis flow meter. The research methodology was based on the use of analytical research methods; a systematic analysis of mechanical and hydrodynamic processes and their mutual influence was performed. Models of Coriolis flow meters were analyzed in order to establish the existing assumptions and assumptions for optimizing the modeling process. Results. A simulation U-shaped model of a Coriolis flow meter has been developed, in which the wear of the walls is uneven, compared to straight-tube structures. The model made it possible to take into account the effect of abrasive wear of the flow tube wall on the accuracy of determining the density and velocity of the liquid at various values of mass flow. Conclusions. Based on the results of simulation modeling, recommendations are proposed for correcting the results of measurements of mass flow of liquids with abrasive elements.
Background. The current level of development of methods for calculating membranes for pressure sensors and their inherent disadvantages is considered. It is shown that the existing calculation methods are mainly focused on metal membranes and do not take into account the influence of membrane parameters on its bending, since they are based on the results of numerical modeling. The aim of the work is to determine the analytical dependence of the deflection of a round flat membrane on the applied pressure and design parameters. Materials and methods. This paper is investigated the equation for the elastic characteristic of the membrane i and its exact analytical solution, describing the dependence of the deflection of the center of a flat circular membrane on the applied pressure, material properties and geometric dimensions of the membrane. Results. The results of numerical modeling of the pressure dependence of the deflection of the center of a silicon membrane at various thicknesses are presented. The deflection values found are significantly lower than the values obtained based on approximate linear dependencies. Conclusions. The dependence of the deflection of the membrane center on pressure, determined on the basis of the obtained accurate analytical expression, is nonlinear, while the error of nonlinearity is insignificant, however, the values of the membrane thickness do not allow for high sensitivity, since the safety margin is several orders of magnitude. This requires further research in order to develop a technique for determining the membrane thickness that ensures the maximum allowable error of non-linearity and maximum sensitivity.
Background. Tin dioxide (SnO2) is a key material for modern opto- and microelectronics. The variety of its synthesis and modification methods, as well as techniques for studying properties, requires systematization to optimize research tasks. The purpose of this work is a comprehensive analysis and classification of modern methods for obtaining and characterizing SnO2-based semiconductor structures. Materials and methods. A critical analysis of the scientific literature on vacuum (magnetron sputtering, atomic layer deposition, chemical vapor deposition) and liquid-phase (sol-gel, spray pyrolysis) synthesis methods, modification methods (doping, plasma treatment), as well as methods for studying structural, morphological, electronic and optical properties was carried out. Results. Based on the analysis, a comprehensive classification table has been developed that systematizes the methods of synthesis, modification and investigation of SnO2 structures, indicating their key parameters, advantages, limitations and target applications. The table allows researchers to make informed choices of methods depending on the required material parameters and final application. Conclusions. The systematization demonstrates a clear correlation between the chosen synthesis method, process parameters and functional characteristics of the obtained SnO2 structures. The developed classifier serves as a practical tool for planning research and developing devices based on tin dioxide.
Background . The aim of this work is to enhance automated diagnostic processes and improve the accuracy of early screening for Alzheimer's disease using machine learning methods. Materials and methods . The data used in the study was obtained using an algorithm developed by the authors for extracting classification features from electroencephalographic signals. The electroencephalographic signals from which the features were extracted were taken from the open-access OpenNeuro database. Methods and models such as neural network and ensemble models, support vector machine models, as well as discriminant analysis and logistic regression, have been investigated by evaluating the classification quality based on data extracted from electroencephalographic signals using several classification quality criteria. Results . Efficiency of various machine learning algorithms was evaluated and the most effective models for the given sets of predictors used in determining the presence of the disease were determined. The results showed high values for the total proportion of correct answers (92.5 %), accuracy (93.3 %), sensitivity (94.4 %), the harmonic mean between sensitivity and accuracy (93.2 %), and the area under the error curve (0.98). Conclusions . The high efficiency of neural network models, support vector machine models, and ensemble and discriminant models in diagnosing Alzheimer's disease using electroencephalographic data has been experimentally confirmed, providing increased accessibility for screening one of the most common and disabling neurodegenerative diseases. Based on the measured indicators, it is recommended to use neural network models ( AUC = 0.98) and support vector machine models ( AUC = 0.97).
Background. The relevance of the research is due to the development of personal aircraft, such as hoverbikes, where the key problem is to ensure flight stability and vibration suppression of the propeller group, which negatively affect on-board optoelectronic systems. The study was conducted on a multirotor apparatus with brushless motors and electronic speed controls. Materials and methods. A cross-correlation function was used to model the relationship between vibrations and image displacement. A set of algorithms was used: adaptive filtering of screw rotation frequencies, median and Gaussian filtering, optical flow analysis (Lucas - Canada) and stabilization based on IMU data. The operational characteristics of the hoverbike are determined by the efficiency of the propeller group and the control system. Results and conclusions. The developed vibration compensation methods increase the noise immunity of optical systems, and the use of brushless motors and electronic speed controls is a technologically sound solution to ensure the required flight qualities.
Background . The operational accuracy and performance of electromechanical measuring and signaling devices are directly determined by the quality of the sensing element. These elements are often flat membranes. Assessing their stress-strain state and fatigue life under a symmetrical loading cycle using analytical relationships is straightforward. However, under asymmetric loading conditions, such as those exposed to destabilizing factors, their use is challenging. A significant number of electromechanical devices in use were developed without the use of automated design and modeling tools. Therefore, their design is currently being critically reviewed for modernization. An example of such a device is a coolant leak alarm for a nuclear power plant reactor. It operates under vibration and high temperatures, and its sensing element, a flat membrane, withstands alternating asymmetric loads. Therefore, the goal of this work is to modernize the leak alarm to increase the fatigue life of the sensing element and, consequently, the reliability and service life of the device as a whole. Materials and methods . Simulation was adopted as the primary research method in the study. A 3D model of the alarm was developed in SolidWorks, and the stress-strain state and fatigue life of the membrane before and after the upgrade were assessed using the Simulation module and finite element method. Results . An analysis of the membrane's stress-strain behavior during operation helped identify and address design flaws. The upgrade resulted in a 30 % increase in the sensing element's fatigue life under maximum loads. Conclusions . The approach, using the example of a leak detector, using the finite element method, allows for the reasonable assignment of design parameters for the product, ensuring the optimal stress-strain state of sensitive elements in the form of flat membranes during operation, thereby predicting and increasing the durability of the entire product.
Background. The object of the study is the cylindrical detail «spindle», for which the requirements for deviation from the roundness of the inner cylindrical surface are established. The subject of the research is methods for measuring deviations from roundness using a coordinate measuring machine. The aim of the work is to increase the accuracy of measuring shape deviation using the coordinate method. Materials and methods. The paper considers methods of basing a cylindrical part on a coordinate measuring machine. Results. The values of the methodological error of measuring the deviation from roundness for various methods of basing the part are obtained. Conclusions. The optimal method of basing a cylindrical detail on a coordinate measuring machine for measuring deviations from roundness is proposed. If there are possible limitations for the optimal basing of the part, alternative ways of installing it on measuring equipment are possible.
Background. The article is devoted to the study of the amplitude-frequency characteristics of physical quantity converters under manufacturing conditions. The relevance of the presented materials lies in the fact that it is during the manufacturing and adjustment of physical quantity converters that their stability and quality of operation at the measurement object are determined. The purpose of the article is to develop control methods and select metrological control tools, in particular, amplitude-frequency and vibration control tools, to ensure the correct measurement of fastflowing processes and minimize the influence of vibration during bench and test site measurements of the parameters of various products and units of rocket-space and flight equipment. Materials and methods. The article provides a description of the designs of modern transducers and explains the proposed methods for monitoring certain metrological characteristics that are very important for the rocket and space industries and aviation. The article also discusses and analyzes the physical responses of transducers to vibration, air and gas flows, and the locations where measuring transducers are installed. Results and conclusions. As a result of the conducted research in the field of monitoring and generating acoustic signals of various amplitudes and frequencies, the correctly achievable amplitude and frequency ranges of the studied measuring transducers were determined, which allowed us to determine the nomenclature of sensors and the methods of their research for exposure to various external interferences.
Background. The simulation method allows for the application of optimal design and engineering solutions in sensor development at the prototyping stage. The objective is to simulate a sensitive element, membrane, and beam elastic element under nominal and overload pressure, taking into account the ambient temperature. Materials and methods. The simulation method employed involves replacing the developed sensor with its model. Solidworks software was employed in the research, allowing for the prevention of errors at the early stages of development. Results and conclusions. Simulation modeling of membrane and beam sensitive elements allowed for the determination of relative radial deformation values under nominal pressure. By approximating the beam SE simulation results, polynomial and power-law dependencies were obtained for determining the relative radial deformation value and the natural frequency depending on the beam thickness. Beam thicknesses were determined for which the safety factor for plastic deformations is 1.3.
Background. The aim of the work is to analyze the methods and means for assessing the biochemical parameters of oral fluid. Materials and methods. The review examines the results of studies published from 2014 to 2024. Results. To systematize the research results, the indicators were divided into the following groups: proteins and peptides, cytokines and inflammation mediators, enzymes and their inhibitors. Conclusions. As a result of the analysis, the most effective and accurate methods for diagnosing inflammatory diseases of the oral cavity for specific diseases were identified to reduce treatment times and reduce the risk of complications.
Background. The development and content of this process of "digitalization" depends on many factors, but the main ones, in our opinion, are determined by the technological level of development of microelectronics, which is the material basis of this process and the perfection of measuring instruments that define the limits of accuracy of technological processes. Materials and methods. Applied examples of applied phase methods in monitoring the characteristics of various electronic devices are considered in the paper. The paper discusses the advantages of the presented phase measurement schemes, the potential capabilities of which can be expanded by their various sensors and auxiliary devices. Results and conclusions. It has been shown that the range of application of phase measurements can be significantly expanded if supplemented with frequency and amplitude measurements.
Background. The problem of automated quality control of colonoscopic studies based on video data analysis is considered. To solve it, it is proposed to detect frames containing images of the dome of the cecum and colon neoplasms in the video stream. Materials and methods. To solve the problem, an integrated approach is proposed that combines modern methods of computer vision and deep machine learning. The solution is based on the YOLOv8 neural network architecture for object detection, supplemented by the Horn – Schunck optical flow algorithm for analyzing the spatio-temporal characteristics of video sequences. Results. The testing showed high efficiency of the proposed method in processing real clinical video data. The algorithm demonstrated stable operation with IoU metric values of up to 0.38, which confirms its applicability for determining areas of interest. Conclusions. The research will serve as a basis for constructing a video stream analysis module in a real endoscopic system based on the developed algorithm.
