All issues

Although the era of exponential performance growth in computer chips has ended, processor core numbers have reached 16 or more even in general-purpose desktop CPUs. As DRAM throughput is unable to keep pace with this computing power growth, CPU designers need to find ways of lowering memory traffic per instruction. The straightforward way to do this is to reduce the miss rate of the last-level cache. Assuming “non-inclusive cache, inclusive directory” (NCID) scheme already implemented, three ways of reducing the cache miss rate further were studied.

The first is to achieve more uniform usage of cache banks and sets by employing hash-based interleaving and indexing. In the experiments in SPEC CPU2017 refrate tests, even the simplest XOR-based hash functions demonstrated a performance increase of 3.2%, 9.1%, and 8.2% for CPU configurations with 16, 32, and 64 cores and last-level cache banks, comparable to the results of more complex matrix-, division- and CRC-based functions.

The second optimisation is aimed at reducing replication at different cache levels by means of automatically switching to the exclusive scheme when it appears optimal. A known scheme of this type, FLEXclusion, was modified for use in NCID caches and showed an average performance gain of 3.8%, 5.4 %, and 7.9% for 16-, 32-, and 64-core configurations.

The third optimisation is to increase the effective cache capacity using compression. The compression rate of the inexpensive and fast BDI*-HL (Base-Delta-Immediate Modified, Half-Line) algorithm, designed for NCID, was measured, and the respective increase in cache capacity yielded roughly 1% of the average performance increase.

All three optimisations can be combined and demonstrated a performance gain of 7.7%, 16% and 19% for CPU configurations with 16, 32, and 64 cores and banks, respectively.

The influence of the process of initiating a rapid local heat release near surface streamlined by supersonic gas (air) flow on the separation region that occurs during a fast turn of the flow was investigated. This surface consists of two planes that form obtuse angle when crossing, so that when flowing around the formed surface, the supersonic gas flow turns by a positive angle, which forms an oblique shock wave that interacts with the boundary layer and causes flow separation. Rapid local heating of the gas above the streamlined surface simulates long spark discharge of submicrosecond duration that crosses the flow. The gas heated in the discharge zone interacts with the separation region. The flow can be considered two-dimensional, so the numerical simulation is carried out in a two-dimensional formulation. Numerical simulation was carried out for laminar regime of flow using the sonicFoam solver of the OpenFOAM software package.

The paper describes a method for constructing a two-dimensional computational grid using hexagonal cells. A study of grid convergence has been carried out. A technique is given for setting the initial profiles of the flow parameters at the entrance to the computational domain, which makes it possible to reduce the computation time by reducing the number of computational cells. A method for non-stationary simulation of the process of rapid local heating of a gas is described, which consists in superimposing additional fields of increased pressure and temperature values calculated from the amount of energy deposited in oncoming supersonic gas flow on the corresponding fields of values obtained in the stationary case. The parameters of the energy input into the flow corresponding to the parameters of the electric discharge process, as well as the parameters of the oncoming flow, are close to the experimental values.

During analyzing numerical simulation data it was found that the initiation of rapid local heating leads to the appearance of a gas-dynamic perturbation (a quasi-cylindrical shock wave and an unsteady swirling flow), which, when interacting with the separation region, leads to a displacement of the separation point downstream. The paper considers the question of the influence of the energy spent on local heating of the gas, and of the position on the streamlined surface of the place of heating relative to the separation point, on the value of its maximum displacement.

The study is based on a three-dimensional hydrodynamic global climate coupled model, including ocean model with real depths and continents configuration, sea ice evolution model and energy and moisture balance atmosphere model. Aerosol concentration from the year 2010 to 2100 is calculated as a controlling parameter to stabilize mean year surface air temperature. It is shown that by this way it is impossible to achieve the space and seasonal uniform approximation to the existing climate, although it is possible significantly reduce the greenhouse warming effect. Climate will be colder at 0.1–0.2 degrees in the low and mid-latitudes and at high latitudes it will be warmer at 0.2–1.2 degrees. The Pareto frontier is investigated and visualized for two parameters — atmospheric temperature mean square deviation for the winter and summer seasons. The Pareto optimal amount of sulfur emissions would be between 23.5 and 26.5 TgS/year.

In this paper we consider a new modification of the discrete version of Mikhailov’s ‘power–society’ model, previously proposed by the author. This modification includes social-economical dynamics and corruption of the system similarly to continuous ‘power–society–economics–corruption’ model but is based on a stochastic cellular automaton describing the dynamics of power distribution in a hierarchy. This new version is founded on previously proposed ‘power–society’ system modeling cellular automaton, its cell state space enriched with variables corresponding to population, economic production, production assets volume and corruption level. The social-economical structure of the model is inherited from Solow and deterministic continuous ‘power–society–economics–corruption’ models. At the same time the new model is flexible, allowing to consider regional differentiation in all social and economical dynamics parameters, to use various production and demography models and to account for goods transit between the regions. A simulation system was built, including three power hierarchy levels, five regions and 100 municipalities. and a number of numerical experiments were carried out. This research yielded results showing specific changes of the dynamics in power distribution in hierarchy when corruption level increases. While corruption is zero (similar to the previous version of the model) the power distribution in hierarchy asymptotically tends to one of stationary states. If the corruption level increases substantially, volume of power in the system is subjected to irregular oscillations, and only much later tends to a stationary value. The meaning of these results can be interpreted as the fact that the stability of power hierarchy decreases when corruption level goes up.

The article deals with the process of creating an effective algorithm for determining the amount of emitted quanta from an X-ray tube in computer tomography (CT) studies. An analysis of domestic and foreign literature showed that most of the work in the field of radiometry and radiography takes the tabulated values of X-ray absorption coefficients into account, while individual dose factors are not taken into account at all since many studies are lacking the Dose Report. Instead, an average value is used to simplify the calculation of statistics. In this regard, it was decided to develop a method to detect the amount of ionizing quanta by analyzing the noise of CT data. As the basis of the algorithm, we used Poisson and Gauss distribution mathematical model of owns’ design of logarithmic value. The resulting mathematical model was tested on the CT data of a calibration phantom consisting of three plastic cylinders filled with water, the X-ray absorption coefficient of which is known from the table values. The data were obtained from several CT devices from different manufacturers (Siemens, Toshiba, GE, Phillips). The developed algorithm made it possible to calculate the number of emitted X-ray quanta per unit time. These data, taking into account the noise level and the radiuses of the cylinders, were converted to X-ray absorption values, after which a comparison was made with tabulated values. As a result of this operation, the algorithm used with CT data of various configurations, experimental data were obtained, consistent with the theoretical part and the mathematical model. The results showed good accuracy of the algorithm and mathematical apparatus, which shows reliability of the obtained data. This mathematical model is already used in the noise reduction program of the CT of own design, where it participates as a method of creating a dynamic threshold of noise reduction. At the moment, the algorithm is being processed to work with real data from computer tomography of patients.

The article is devoted to the construction and investigation of an averaged mathematical model of an aluminum-cobalt-molybdenum hydrocracking catalyst oxidative regeneration. The oxidative regeneration is an effective means of restoring the activity of the catalyst when its granules are coating with coke scurf.

The mathematical model of this process is a nonlinear system of ordinary differential equations, which includes kinetic equations for reagents’ concentrations and equations for changes in the temperature of the catalyst granule and the reaction mixture as a result of isothermal reactions and heat transfer between the gas and the catalyst layer. Due to the heterogeneity of the oxidative regeneration process, some of the equations differ from the standard kinetic ones and are based on empirical data. The article discusses the scheme of chemical interaction in the regeneration process, which the material balance equations are compiled on the basis of. It reflects the direct interaction of coke and oxygen, taking into account the degree of coverage of the coke granule with carbon-hydrogen and carbon-oxygen complexes, the release of carbon monoxide and carbon dioxide during combustion, as well as the release of oxygen and hydrogen inside the catalyst granule. The change of the radius and, consequently, the surface area of coke pellets is taken into account. The adequacy of the developed averaged model is confirmed by an analysis of the dynamics of the concentrations of substances and temperature.

The article presents a numerical experiment for a mathematical model of oxidative regeneration of an aluminum-cobalt-molybdenum hydrocracking catalyst. The experiment was carried out using the Kutta–Merson method. This method belongs to the methods of the Runge–Kutta family, but is designed to solve stiff systems of ordinary differential equations. The results of a computational experiment are visualized.

The paper presents the dynamics of the concentrations of substances involved in the oxidative regeneration process. A conclusion on the adequacy of the constructed mathematical model is drawn on the basis of the correspondence of the obtained results to physicochemical laws. The heating of the catalyst granule and the release of carbon monoxide with a change in the radius of the granule for various degrees of initial coking are analyzed. There are a description of the results.

In conclusion, the main results and examples of problems which can be solved using the developed mathematical model are noted.

The paper presents agent-based model for city formation named Technoscape which is both local and nonlocal. Technoscape can, to a certain degree, be also assumed as a model for emergence of global economy. The current version of the model implements very simple way of agents’ behavior and interaction, still the model provides rather interesting spatio-temporal patterns.

Locality and non-locality mean here the spatial features of the way the agents interact with each other and with geographical space upon which the evolution takes place. Technoscape agent is some conventional artisan, family, or а producing and trading firm, while there is no difference between production and trade. Agents are located upon and move through bounded two-dimensional space divided into square cells. The model demonstrates processes of agents’ concentration in a small set of cells, which is interpreted as «city» formation. Agents are immortal, they don’t mutate and evolve, though this is interesting perspective for the evolution of the model itself.

Technoscape provides some distinctively new type of self-organization. Partially, this type of selforganization resembles the behavior of segregation model by Thomas Shelling, still that model has evolution rules substantially different from Technoscape. In Shelling model there exist avalanches still simple equilibria exist if no new agents are added to the game board, while in Technoscape no such equilibria exist. At best, we can observe quasi-equilibrium, slowly changing global states.

One non-trivial phenomenon Technoscape exhibits, which also contrasts to Shelling segregation model, is the ability of agents to concentrate in local cells (interpreted as cities) even explicitly and totally ignoring local interactions, using non-local interactions only.

At the same time, while the agents tend to concentrate in large one-cell cities, large scale of such cities does not guarantee them from decay: there always exists a process of «enticement» of agents and their flow to new cities.

This article is devoted to the development of an algorithm for trajectory control of a highly maneuverable four-wheeled robotic transport platform equipped with mecanum wheels, in order to organize its movement behind some moving object. The calculation of the kinematic ratios of this platform in a fixed coordinate system is presented, which is necessary to determine the angular velocities of the robot wheels depending on a given velocity vector. An algorithm has been developed for the robot to follow a mobile object on a plane without obstacles based on the use of a modified chase method using different types of control functions. The chase method consists in the fact that the velocity vector of the geometric center of the platform is co-directed with the vector connecting the geometric center of the platform and the moving object. Two types of control functions are implemented: piecewise and constant. The piecewise function means control with switching modes depending on the distance from the robot to the target. The main feature of the piecewise function is a smooth change in the robot’s speed. Also, the control functions are divided according to the nature of behavior when the robot approaches the target. When using one of the piecewise functions, the robot’s movement slows down when a certain distance between the robot and the target is reached and stops completely at a critical distance. Another type of behavior when approaching the target is to change the direction of the velocity vector to the opposite, if the distance between the platform and the object is the minimum allowable, which avoids collisions when the target moves in the direction of the robot. This type of behavior when approaching the goal is implemented for a piecewise and constant function. Numerical simulation of the robot control algorithm for various control functions in the task of chasing a target, where the target moves in a circle, is performed. The pseudocode of the control algorithm and control functions is presented. Graphs of the robot’s trajectory when moving behind the target, speed changes, changes in the angular velocities of the wheels from time to time for various control functions are shown.

This work provides a new approach for constructing high-precision routes based on data from transport detectors inside the SUMO traffic modeling package. Existing tools such as flowrouter and routeSampler have a number of disadvantages, such as the lack of interaction with the network in the process of building routes. Our rlRouter uses multi-agent reinforcement learning (MARL), where the agents are incoming lanes and the environment is the road network. By performing actions to launch vehicles, agents receive a reward for matching data from transport detectors. Parameter Sharing DQN with the LSTM backbone of the Q-function was used as an algorithm for multi-agent reinforcement learning.

Since the rlRouter is trained inside the SUMO simulation, it can restore routes better by taking into account the interaction of vehicles within the network with each other and with the network infrastructure. We have modeled diverse traffic situations on three different junctions in order to compare the performance of SUMO’s routers with the rlRouter. We used Mean Absoluter Error (MAE) as the measure of the deviation from both cumulative detectors and routes data. The rlRouter achieved the highest compliance with the data from the detectors. We also found that by maximizing the reward for matching detectors, the resulting routes also get closer to the real ones. Despite the fact that the routes recovered using rlRouter are superior to the routes obtained using SUMO tools, they do not fully correspond to the real ones, due to the natural limitations of induction-loop detectors. To achieve more plausible routes, it is necessary to equip junctions with other types of transport counters, for example, camera detectors.

Simulation model of corruption in hierarchical systems which takes into account individual strategies of elements and collective behavior of large groups is proposed. Evolution of various characteristics like level of corruption or ratio of corrupted elements and their dependence on external parameters are discussed. The effectiveness of various anticorruptional strategies is examined by means of numeric analysis.