All issues

In this paper we consider the controlled motion of a helical body with three blades in an ideal fluid, which is executed by rotating three internal rotors. We set the problem of selecting control actions, which ensure the motion of the body near the predetermined trajectory. To determine controls that guarantee motion near the given curve, we propose methods based on the application of hybrid genetic algorithms (genetic algorithms with real encoding and with additional learning of the leader of the population by a gradient method) and artificial neural networks. The correctness of the operation of the proposed numerical methods is estimated using previously obtained differential equations, which define the law of changing the control actions for the predetermined trajectory.

In the approach based on hybrid genetic algorithms, the initial problem of minimizing the integral functional reduces to minimizing the function of many variables. The given time interval is broken up into small elements, on each of which the control actions are approximated by Lagrangian polynomials of order 2 and 3. When appropriately adjusted, the hybrid genetic algorithms reproduce a solution close to exact. However, the cost of calculation of 1 second of the physical process is about 300 seconds of processor time.

To increase the speed of calculation of control actions, we propose an algorithm based on artificial neural networks. As the input signal the neural network takes the components of the required displacement vector. The node values of the Lagrangian polynomials which approximately describe the control actions return as output signals . The neural network is taught by the well-known back-propagation method. The learning sample is generated using the approach based on hybrid genetic algorithms. The calculation of 1 second of the physical process by means of the neural network requires about 0.004 seconds of processor time, that is, 6 orders faster than the hybrid genetic algorithm. The control calculated by means of the artificial neural network differs from exact control. However, in spite of this difference, it ensures that the predetermined trajectory is followed exactly.

This work presents the model of heat wall functions FlowVision (WFFV), which allows simulation of nonisothermal flows of fluid and gas near solid surfaces on relatively coarse grids with use of turbulence models. The work follows the research on the development of wall functions applicable in wide range of the values of quantity y+. Model WFFV assumes smooth profiles of the tangential component of velocity, turbulent viscosity, temperature, and turbulent heat conductivity near a solid surface. Possibility of using a simple algebraic model for calculation of variable turbulent Prandtl number is investigated in this study (the turbulent Prandtl number enters model WFFV as parameter). The results are satisfactory. The details of implementation of model WFFV in the FlowVision software are explained. In particular, the boundary condition for the energy equation used in high-Reynolds number calculations of non-isothermal flows is considered. The boundary condition is deduced for the energy equation written via thermodynamic enthalpy and via full enthalpy. The capability of the model is demonstrated on two test problems: flow of incompressible fluid past a plate and supersonic flow of gas past a plate (M = 3).

Analysis of literature shows that there exists essential ambiguity in experimental data and, as a consequence, in empirical correlations for the Stanton number (that being a dimensionless heat flux). The calculations suggest that the default values of the model parameters, automatically specified in the program, allow calculations of heat fluxes at extended solid surfaces with engineering accuracy. At the same time, it is obvious that one cannot invent universal wall functions. For this reason, the controls of model WFFV are made accessible from the FlowVision interface. When it is necessary, a user can tune the model for simulation of the required type of flow.

The proposed model of wall functions is compatible with all the turbulence models implemented in the FlowVision software: the algebraic model of Smagorinsky, the Spalart-Allmaras model, the SST $k-\omega$ model, the standard $k-\varepsilon$ model, the $k-\varepsilon$ model of Abe, Kondoh, Nagano, the quadratic $k-\varepsilon$ model, and $k-\varepsilon$ model FlowVision.

The mathematical model, finite-difference schemes and algorithms for computation of transient thermoand hydrodynamic processes involved in commissioning the unified system including the oil producing well, electrical submersible pump and fractured-porous reservoir with bottom water are developed. These models are implemented in the computer package to simulate transient processes with simultaneous visualization of their results along with computations. An important feature of the package Oil-RWP is its interaction with the special external program GCS which simulates the work of the surface electric control station and data exchange between these two programs. The package Oil-RWP sends telemetry data and current parameters of the operating submersible unit to the program module GCS (direct coupling). The station controller analyzes incoming data and generates the required control parameters for the submersible pump. These parameters are sent to Oil-RWP (feedback). Such an approach allows us to consider the developed software as the “Intellectual Well System”.

Some principal results of the simulations can be briefly presented as follows. The transient time between inaction and quasi-steady operation of the producing well depends on the well stream watering, filtration and capacitive parameters of oil reservoir, physical-chemical properties of phases and technical characteristics of the submersible unit. For the large time solution of the nonstationary equations governing the nonsteady processes is practically identical to the inverse quasi-stationary problem solution with the same initial data. The developed software package is an effective tool for analysis, forecast and optimization of the exploiting parameters of the unified oil-producing complex during its commissioning into the operating regime.

Certifying a transport airplane for the flights under icing conditions requires calculations aimed at definition of the dimensions and shapes of the ice bodies formed on the airplane surfaces. Up to date, software developed in Russia for simulation of ice accretion, which would be authorized by Russian certifying supervisory authority, is absent. This paper describes methodology IceVision recently developed in Russia on the basis of software FlowVision for calculations of ice accretion on airplane surfaces.

The main difference of methodology IceVision from the other approaches, known from literature, consists in using technology Volume Of Fluid (VOF — volume of fluid in cell) for tracking the surface of growing ice body. The methodology assumes solving a time-depended problem of continuous grows of ice body in the Euler formulation. The ice is explicitly present in the computational domain. The energy equation is integrated inside the ice body. In the other approaches, changing the ice shape is taken into account by means of modifying the aerodynamic surface and using Lagrangian mesh. In doing so, the heat transfer into ice is allowed for by an empirical model.

The implemented mathematical model provides capability to simulate formation of rime (dry) and glaze (wet) ice. It automatically identifies zones of rime and glaze ice. In a rime (dry) ice zone, the temperature of the contact surface between air and ice is calculated with account of ice sublimation and heat conduction inside the ice. In a glaze (wet) ice zone, the flow of the water film over the ice surface is allowed for. The film freezes due to evaporation and heat transfer inside the air and the ice. Methodology IceVision allows for separation of the film. For simulation of the two-phase flow of the air and droplets, a multi-speed model is used within the Euler approach. Methodology IceVision allows for size distribution of droplets. The computational algorithm takes account of essentially different time scales for the physical processes proceeding in the course of ice accretion, viz., air-droplets flow, water flow, and ice growth. Numerical solutions of validation test problems demonstrate efficiency of methodology IceVision and reliability of FlowVision results.

A mathematical model describing the irreversible processes of polarization and deformation of polycrystalline ferroelectrics in external electric and mechanical fields of high intensity is presented, as a result of which the internal structure changes and the properties of the material change. Irreversible phenomena are modeled in a three-dimensional setting for the case of simultaneous action of an electric field and mechanical stresses. The object of the research is a representative volume in which the residual phenomena in the form of the induced and irreversible parts of the polarization vector and the strain tensor are investigated. The main task of modeling is to construct constitutive relations connecting the polarization vector and strain tensor, on the one hand, and the electric field vector and mechanical stress tensor, on the other hand. A general case is considered when the direction of the electric field may not coincide with any of the main directions of the tensor of mechanical stresses. For reversible components, the constitutive relations are constructed in the form of linear tensor equations, in which the modules of elasticity and dielectric permeability depend on the residual strain, and the piezoelectric modules depend on the residual polarization. The constitutive relations for irreversible parts are constructed in several stages. First, an auxiliary model was constructed for the ideal or unhysteretic case, when all vectors of spontaneous polarization can rotate in the fields of external forces without mutual influence on each other. A numerical method is proposed for calculating the resulting values of the maximum possible polarization and deformation values of an ideal case in the form of surface integrals over the unit sphere with the distribution density obtained from the statistical Boltzmann law. After that the estimates of the energy costs required for breaking down the mechanisms holding the domain walls are made, and the work of external fields in real and ideal cases is calculated. On the basis of this, the energy balance was derived and the constitutive relations for irreversible components in the form of equations in differentials were obtained. A scheme for the numerical solution of these equations has been developed to determine the current values of the irreversible required characteristics in the given electrical and mechanical fields. For cyclic loads, dielectric, deformation and piezoelectric hysteresis curves are plotted.

The developed model can be implanted into a finite element complex for calculating inhomogeneous residual polarization and deformation fields with subsequent determination of the physical modules of inhomogeneously polarized ceramics as a locally anisotropic body.

In this paper, we consider convex stochastic optimization problems arising in machine learning applications (e. g., risk minimization) and mathematical statistics (e. g., maximum likelihood estimation). There are two main approaches to solve such kinds of problems, namely the Stochastic Approximation approach (online approach) and the Sample Average Approximation approach, also known as the Monte Carlo approach, (offline approach). In the offline approach, the problem is replaced by its empirical counterpart (the empirical risk minimization problem). The natural question is how to define the problem sample size, i. e., how many realizations should be sampled so that the quite accurate solution of the empirical problem be the solution of the original problem with the desired precision. This issue is one of the main issues in modern machine learning and optimization. In the last decade, a lot of significant advances were made in these areas to solve convex stochastic optimization problems on the Euclidean balls (or the whole space). In this work, we are based on these advances and study the case of arbitrary balls in the $p$-norms. We also explore the question of how the parameter $p$ affects the estimates of the required number of terms as a function of empirical risk.

In this paper, both convex and saddle point optimization problems are considered. For strongly convex problems, the existing results on the same sample sizes in both approaches (online and offline) were generalized to arbitrary norms. Moreover, it was shown that the strong convexity condition can be weakened: the obtained results are valid for functions satisfying the quadratic growth condition. In the case when this condition is not met, it is proposed to use the regularization of the original problem in an arbitrary norm. In contradistinction to convex problems, saddle point problems are much less studied. For saddle point problems, the sample size was obtained under the condition of $\gamma$-growth of the objective function. When $\gamma = 1$, this condition is the condition of sharp minimum in convex problems. In this article, it was shown that the sample size in the case of a sharp minimum is almost independent of the desired accuracy of the solution of the original problem.

This article discusses the mechanical properties of carbon nanotube (CNT)-reinforced platinum under uniaxial tensile loading using the molecular dynamics method. A review of current computational and experimental studies on the use of carbon nanotube-reinforced composites from a structural point of view. However, quantitative and qualitative studies of CNTs to improve the properties of composites are still rare. Composite selection is a promising application for platinum alloys in many cases where they may be subjected to mechanical stress, including in biocompatibility sources. Pt-reinforced with CNTs may have additional possibilities for implantation of the implant and at the same time obtain the required mechanical characteristics.

The structure of the composite is composed of a Pt crystal with a face-centered cubic lattice with a constant of 3.92 Å and a carbon nanotube. The Pt matrix has the shape of a cube with dimensions of $43.1541 Å \times 43.1541 Å \times 43.1541 Å$. The hole size in the average platinum dimension is the radius of the carbon nanotube of the «zigzag» type (8,0), which is 2.6 Å. A carbon nanotube is placed in a hole with a radius of 4.2 Å. At such parameters, the maximum energy level was mutually observed. The model under consideration is contained in 320 atomic bombs and 5181 atomic platinum. The volume fraction of deaths in the Pt-C composite is 5.8%. At the first stage of the study, the strain rate was analyzed for stress-strain and energy change during uniaxial action on the Pt-C composite.

Analysis of the strain rate study showed that the consumption yield strength increases with high strain rate, and the elasticity has increased density with decreasing strain rate. This work also increased by 40% for Pt-C, the elasticity of the composite decreased by 42.3%. In general, fracture processes are considered in detail, including plastic deformation on an atomistic scale.

This paper considers various approaches to averaging the generalized travel costs calculated for different modes of travel in the transportation network. The mode of transportation is understood to mean both the mode of transport, for example, a car or public transport, and movement without the use of transport, for example, on foot. The task of calculating the trip matrices includes the task of calculating the total matrices, in other words, estimating the total demand for movements by all modes, as well as the task of splitting the matrices according to the mode, also called modal splitting. To calculate trip matrices, gravitational, entropy and other models are used, in which the probability of movement between zones is estimated based on a certain measure of the distance of these zones from each other. Usually, the generalized cost of moving along the optimal path between zones is used as a distance measure. However, the generalized cost of movement differs for different modes of movement. When calculating the total trip matrices, it becomes necessary to average the generalized costs by modes of movement. The averaging procedure is subject to the natural requirement of monotonicity in all arguments. This requirement is not met by some commonly used averaging methods, for example, averaging with weights. The problem of modal splitting is solved by applying the methods of discrete choice theory. In particular, within the framework of the theory of discrete choice, correct methods have been developed for averaging the utility of alternatives that are monotonic in all arguments. The authors propose some adaptation of the methods of the theory of discrete choice for application to the calculation of the average cost of movements in the gravitational and entropy models. The transfer of averaging formulas from the context of the modal splitting model to the trip matrix calculation model requires the introduction of new parameters and the derivation of conditions for the possible value of these parameters, which was done in this article. The issues of recalibration of the gravitational function, which is necessary when switching to a new averaging method, if the existing function is calibrated taking into account the use of the weighted average cost, were also considered. The proposed methods were implemented on the example of a small fragment of the transport network. The results of calculations are presented, demonstrating the advantage of the proposed methods.

The development of a high-temperature gas-cooled reactor (HTGR) constituting a part of nuclear power-and-process station and intended for large-scale hydrogen production is now in progress in the Russian Federation. One of the key objectives in development of the high-temperature gas-cooled reactor is the computational justification of the accepted design.

The article gives the procedure for the computational analysis of thermal and physical characteristics of the high-temperature gas-cooled reactor. The procedure is based on the use of the state-of-the-art codes for personal computer (PC).

The objective of thermal and physical analysis of the reactor as a whole and of the core in particular was achieved in three stages. The idea of the first stage is to justify the neutron physical characteristics of the block-type core during burn-up with the use of the MCU-HTR code based on the Monte Carlo method. The second and the third stages are intended to study the coolant flow and the temperature condition of the reactor and the core in 3D with the required degree of detailing using the FlowVision and the ANSYS codes.

For the purpose of carrying out the analytical studies the computational models of the reactor flow path and the fuel assembly column were developed.

As per the results of the computational modeling the design of the support columns and the neutron physical characteristics of the fuel assembly were optimized. This results in the reduction of the total hydraulic resistance of the reactor and decrease of the maximum temperature of the fuel elements.

The dependency of the maximum fuel temperature on the value of the power peaking factors determined by the arrangement of the absorber rods and of the compacts of burnable absorber in the fuel assembly is demonstrated.

In today’s technology-driven world, lite devices like Internet of Things (IoT) devices and microcontrollers (MCUs) are becoming increasingly common. These devices are more energyefficient and affordable, often with reduced features compared to the standard versions such as very limited memory and processing power for typical machine learning models. However, modern machine learning models can have millions of parameters, resulting in a large memory footprint. This complexity not only makes it difficult to deploy these large models on resource constrained devices but also increases the risk of latency and inefficiency in processing, which is crucial in some cases where real-time responses are required such as autonomous driving and medical diagnostics. In recent years, neural networks have seen significant advancements in model optimization techniques that help deployment and inference on these small devices. This narrative review offers a thorough examination of the progression and latest developments in neural network optimization, focusing on key areas such as quantization, pruning, knowledge distillation, and neural architecture search. It examines how these algorithmic solutions have progressed and how new approaches have improved upon the existing techniques making neural networks more efficient. This review is designed for machine learning researchers, practitioners, and engineers who may be unfamiliar with these methods but wish to explore the available techniques. It highlights ongoing research in optimizing networks for achieving better performance, lowering energy consumption, and enabling faster training times, all of which play an important role in the continued scalability of neural networks. Additionally, it identifies gaps in current research and provides a foundation for future studies, aiming to enhance the applicability and effectiveness of existing optimization strategies.