Vol 33, No 4 (2025)

The study presents work aimed at improving the efficiency of the information, measurement and control system of vibration monitoring of turbine units. A key aspect of the study is the analysis of the effect of vibrations with frequencies not multiple of the reverse frequency of the turbine units on the accuracy of determining the parameters of piezo accelerometers. Piezoaccelerometers act as primary measuring devices in the vibration monitoring system of turbine units, and their correct operation is critically important for reliable monitoring of equipment condition. The authors improved the previously developed method for determining piezoaccelerometer parameters, moving from using a neural network with one hidden layer to using deep neural networks. As part of the study, two modifications of the method were studied: using a recurrent neural network and a direct propagation network. The article describes in detail the architecture and key parameters of these networks, as well as the training parameters and the composition of the training dataset. A series of experiments was conducted, during which the results of training and testing neural networks were analyzed. The maximum allowable amplitude of vibration components with frequencies not multiple of the reverse frequency of the turbine unit has been established, at which the neural network provides the required accuracy in determining the parameters of free oscillations of the piezoelectric accelerometer sensor element and its conversion coefficient. The results obtained using deep neural networks are compared with the data obtained using prototype methods and an artificial neural network with one hidden layer. The study showed that deep neural networks provide higher accuracy in determining parameters. Among the studied types of networks, the direct distribution network has demonstrated the best results, which makes it the most preferable for practical use. In addition, it is noted that the use of deep neural networks reduces the number of neurons in hidden layers, which leads to a reduction in computational operations and resource consumption. The results obtained indicate the possibility of using the proposed method in equipment operating conditions that are more stringent than those in which prototype methods are effective.

The similarity of engine characteristics is an important factor in the controllability and stability of the operation of mini-unmanned aerial vehicles (UAVs). The coordinated operation of engines with different characteristics is ensured by the flight controller, through commands sent to the speed controller of each engine. The selection of commands based on the characteristics of individual engines requires the expenditure of computing resources of the flight controller, which negatively affects the controllability and stability of the mini-UAV. Because of this, the selection of engines with similar characteristics is a factor that increases the controllability and stability of mini-UVs. A method for quantifying the degree of similarity of engine characteristics based on the results of bench tests is proposed, based on solving the inverse problem of determining the distribution law of a random argument function. The method is fully formalized, which makes it possible to implement a software product based on it, which can be used as a functional component in bench testing systems for both individual engines and mini-UVs. Promising bench-testing tasks have been identified, the solution of which is possible based on the developed method.

The increasing complexity of control plants leads to a decrease in the degree of their research with an increasing level of unreliability of a priory information, uncertainty in the formulation of tasks for the synthesis of automatic control systems and forces the expansion of the use of approximate mathematical models. As a result, iterative problem solving methods are more widely used, implementing the principle of consistent approximation to the desired result that meets the specified requirements. Iterative algorithms provide for the cyclic execution of a number of operations that implement the iterative method used to solve the problem. Despite the wide range of software development tools, numerous libraries of standard and other various mathematical programs, the designer does not have the opportunity to independently developing the necessary software to solve the problems of control systems design. This is especially important when using iterative problem solving methods, which involve variability in the source data and algorithms with their cyclic application in the problem solving process, which creates additional difficulties. The use of automatic program synthesis is limited by the algorithmic unresolvable of the cyclic program synthesis problem. To solve this problem, a methodology is proposed for applying a two-stage approach to the automatic solution of declarative set tasks using iterative methods, including the automatic construction of an action plan and its subsequent execution in a generated virtual algorithmic structure, including conditions and cycles. The implementation of the approach is based on the planning of actions by a planning artificial neural network and the execution of the constructed action plan into a generated algorithmic structure controlled by the conditions of applicability and requirements for the results of solving the problem. The resolvability of action planning algorithms is shown, which generate a task solution plan in the form of a subset of the axioms of the knowledge model describing the necessary operations to build the desired result. A generalized scheme of cyclic execution of operations of the task solution plan is proposed, controlled by the requirements for the desired result, presented in the form of aggregated operations, including an analysis of the conditions of applicability. The resolvability of the proposed two-stage method is proved. An example illustrating the application of a two-stage method to solve a model problem is considered. Its practical interpretation is shown, including for control tasks of an indefinite plant with multiplicative frequency uncertainty.

This study investigates the enhancement of thermo-hydraulic performance in compact heat exchangers, which are critical components in aerospace, automotive, and energy systems where size, weight, and thermal efficiency are paramount. While spiral fins and perforated surfaces individually improve heat transfer, their combined application in plate-fin heat exchangers remains largely unexplored. This work introduces a novel perforated spiral fin heat exchanger design, featuring bi-directional spiral fins with D-shaped perforations. This configuration aims to augment heat transfer efficiency by leveraging swirling flow within spiral channels and flow mixing induced by the perforations, while simultaneously mitigating hydraulic losses. Three-dimensional numerical simulations were performed to evaluate the impact of perforation height on performance, using the Colburn j-factor, Fanning friction f-factor, and performance evaluation criterion (JF) as metrics. Compared to a baseline spiral fin design, the proposed perforated spiral fin heat exchanger demonstrated significant improvements. At a Reynolds number of 3000, increasing the perforation height from 0 mm to 1.75 mm resulted in a j-factor enhancement from 8.55% to 14.77%, an f-factor reduction of up to 14.82%, and a JF factor increase ranging from 6.16% to 21.07%. These results confirm that the perforations effectively reduce pressure losses associated with flow separation and vortices while simultaneously increasing heat transfer efficiency. This research provides the first comprehensive analysis of the synergistic effects between spiral-induced swirl and fin perforation in a perforated spiral fin heat exchanger, offering new physical insights into their interaction.