No 4 (2024)

The paper presents the possibility of identifying applications by network traffic. Classical classifier methods are applied for identification. The paper proposes a model for collecting and processing input network data, as well as an algorithm for forming a feature description for classifying network applications to improve the accuracy of application identification. The results of the experiments made it possible to identify the advantages and disadvantages of the classification methods used.

The article discusses the problem of representing a nonlinear dynamic system in the state space using the quasi-linearization method. The developed method for constructing a quasi-linear model is based on the principles of structural-parametric synthesis. The first stage of construction consists in the analytical output of the structure of the quasi-linear model from nonlinear differential equations with the identification of variable parameters, on which the dynamic characteristics of the control object depend. At the stage of parametric synthesis, identification of unknown elements of the resulting structure is carried out in the form of functional dependencies on the corresponding variables. For this, the task of approximation in the form of a linear generalized model using algorithms of sparse regression and heuristic decomposition of this model on the necessary functional dependencies is solved. As an example, the quasi-linearization method is applied to the nonlinear model of the isolated movement of the unmanned aerial vehicle in the longitudinal channel relative to the center of the masses. The proposed method can be applied to other models with similar properties to solve problems of controlling nonlinear dynamic systems with modern methods of control theory.

The article presents the results of calculations of the initial stage of the RIA type emergency mode with a stepwise introduction of positive reactivity using a simplified fuel element model. This process is accompanied by heating of the dispersion fuel composition in the most heat-stressed sections of the fuel elements above the melting point of the silumin matrix. The time delay in the heating of part of the core fuel due to energy losses on the destruction of the crystalline structure of silumin leads to a decrease in the energy transmitted to the coolant and a weakening of the Doppler effect. As a result of these processes, the nuclear energy released during the power «flash» increases with the overcompensation of melting losses, mainly due to thermal inertia. Modeling of the beginning of the model emergency situation shows that the maximum temperatures in the core decreased, but the specific load on the fuel element increased during the «flash» of power.

The relationship between the ratio of the total energy released during the «flash» process and the energy used only to heat the fuel element from the volume fraction of the fuel composition in the core that has reached the melting point of silumin is obtained. It is recommended to take into account the melting process of silumin during studies of significantly beyond design basis reactivity accidents.

The article presents the results of a study that continues a cycle of calculation and experimental work aimed at developing a high-temperature magnetohydrodynamic filter designed for water treatment systems of nuclear power plants. The results of numerical modeling of three design variants of the modernized flow part of the hydrodynamic filter using permanent magnets made of an alloy of iron, boron and neodymium of the N42SH brand are presented. Modeling of hydrodynamic processes in a viscous incompressible fluid flow was carried out in a commercial calculation code using the Reynolds stress model. In this model, turbulent stresses in the fluid flow are nonlinearly related to the velocity components, which leads to a more accurate solution regarding the spatial distribution of velocity within the RANS approach. The optimal value of the magnetic system height, based on the results of the calculation study, was 60 mm. The calculated value of hydraulic resistance at maximum flow rate was 52.94 kPa. The problem of straightening the flow at the outlet section of a magnetohydrodynamic filter in order to minimize the hydraulic resistance of the structure and ensure the most uniform turbulent velocity profile is also investigated. The results of numerical modeling of seven design variants of a flow straightener based on the use of various finning options made of straight plates on the inner generatrix of the channel are presented. The analysis of the results of the calculation study was carried out and the most effective and optimal design of the straightener from the point of view of reliability and ease of manufacture was proposed.

The article analyzes measures to reduce the aerodynamic drag of a light commercial vehicle. A review of existing solutions showed that the most promising method of reducing aerodynamic drag when transition from an airborne version to a version with an awning is to install a fairing on the roof of the vehicle. The effect of the fairing on fuel consumption indicators was assessed. The object of study is the basic model of the GAZ vehicle – A21R23, equipped with a gasoline internal combustion engine. Its fuel consumption is estimated based on the results of tests carried out at constant speeds: 60, 80 and 100 km/h, in the NEDC driving cycle. It has been established that a vehicle with an awning has a 7% increase in fuel consumption. The use of a purchased fairing showed insignificant results in saving fuel consumption, so this led to the development of our own fairing. The obtained computer modeling data proved that installation of the developed fairing can reduced the aerodynamic drag coefficient by 21 %, and fuel savings in the NEDC cycle will be 5 %.

The article presents an algorithm for ensuring a given level of mobility of a pseudo-walking robot with a three-roller propulsion unit at the design stage. The driving performance of the robot under study is assessed on the basis of mathematical simulation of its movement using software packages for studying the dynamics of rigid body systems. The mobility requirements are formulated in the form of obstacles of a virtual test site. The results of modeling the overcoming of these obstacles – the loads acting on the propulsion unit are presented. The parameters of the drives and propulsion unit were selected to ensure that the robot passes all obstacles. The response of the structure to a change in the control signal is determined: acceleration time to maximum speed, braking time to a complete stop, maximum speed of the robot. The robot's deviation from the straight course when moving along a slope was predicted. A picture of the loads and a forecast of the robot’s cross-country ability were obtained, and it was concluded that the robot meets the stated requirements.

The article proposes a method for analyzing the cross-country ability of a mobile robot, which can be used during the design stage. The method involves analyzing the cross-country ability using an iterative algorithm based on analytical equations of rigid body statics, while, as an assumption, the rolling resistance of the wheel is not taken into account. The program is designed to determine the optimal position of the machine's center of mass needed to overcome the obstacles, taking into account the available traction force from the engines. The dependences necessary to find the forces and moments acting when the wheel interacts with the support surface are given. The described algorithm for calculating the forces acting on the machine is not demanding of computing power, unlike modern methods by numerical mathematical modeling using for these tasks. The adequacy of the proposed method for assessing cross-country ability was confirmed using a simple field experiment. The presented approach may be useful for developing a control system for machines capable of changing the location of the center of gravity when moving.