Полоцкий государственный университет

Polotsk
State
University

ESTIMATION METHODS OF SPEECH PROTECTION
V. ZHELEZNYAK, K. RAKHANAU

Considered methods of noise signal and harmonic measurement signal. Pointed to a number of existing advantages of the harmonic signal. Proposed a method for measuring signal linear frequency modulation to reduce the methodical error of estimation of speech intelligibility. Lowered the threshold effect signal linear frequency modulation. Suggest a method estimating the parameters of linear frequency modulation signal with the fine structure of the spectral components of the density distribution of the signal Wigner energy. The method of signal linear frequency modulation takes advantage of the method of harmonic signal and eliminates its disadvantage, caused considerable irregularity of the frequency response channel information leaks.

COMBINED ALGORITHM FOR SOLVING SYSTEM OF NONLINEAR EQUATIONS GASDYNAMIC PROBLEM FOR GAS TRANSPORTATION NETWORKS
D. HLUKHAU, S. AVILKIN

This paper presents an algorithm for solving systems of nonlinear equations with usage a local heuristics, taking into account the features of such systems in models of networks of gas transportation. We analyze results of the comparative tests of the algorithm with classical algorithms such as the method of coordinate relaxation, block relaxation, conjugate gradient method, and, also, features of the selected class of nonlinear systems. We concluded about high sensitivity of algorithms for solving this class of systems from the initial approximation and the parameters of the iterative process. We proposed method for self-optimization for solving the problem for specific networks of gas transmission. Also recommendations about usageof thelocal heuristics in the solution process of the such systems are determined.

INFORMATION INTERPRETATION OF PHYSICAL AND CHEMICAL ANALYSIS, MULTIFRACTAL AND WAVELET ANALYSIS OF STRUCTURE AND FRICTION SURFACE RELIEF
A. KOLMAKOV, A. ZVEREV, M. KHEIFETZ, S. KOUKHTA

Using system approaches provides essential development of traditional methods in materiology and their effective application for objects with difficult multiscale structure and for nano-structural materials as well. It shows application of complex physical and chemical analysis, multifractal and wavelet analysis to describe material structures in qualitative and quantitative way, which allows revealing the differences poorly distinguished with other separate methods and estimating system characteristics of material structures and characterizing the process in self-organizing in structures. Formation by tearing of friction surface structures for composite materials with additives of Al2O3 and SiC is described in comparable thermodynamic conditions. Adding SiC in comparison with Al2O3 promotes to a lesser extent fragmentation and increase in frequency of setting and breaking processes. Adding graphite to reinforced SiC composite material results in increase in nonequilibrium of thermodynamic condition of structure formation. Additional introduction of graphite in even greater degree strengthens distinctions between materials with SiC and Al2O3 additives, which results in formation of the least ordered structures of friction surface.

A FORM COMPLEXITY AND COMPUTER-AIDED TECHNOLOGIES FOR ITS MEASUREMENT
D. SVIRSKY

The problem of determination of complexity of the three-dimensional objects surface form is considered. The main methodological approaches to this characteristic measurement on the basis of the theory of sets, the theory of the information, the theory of algorithms and the theory of calculations are analyzed. Necessity of working out of a special objective indicator of complexity of a geometrical configuration of natural and technical volume forms is proved. The quantitative criterion of complexity of the form on the basis of geometrical characteristics is offered as normalized factor of the three-dimensional object surface form development. The possibilities of the computer technologies for a configuration complexity factor calculation are shown. Influence of the discrete size on accuracy of measurement of complexity of the geometrical form of objects is investigated. The mathematical model corresponding regression dependence is constructed. The directions of the further researches which should be working out are specified. In particular, the techniques of an estimation of a various sort symmetry influence on numerical value of the surface form complexity indicator.

SIMULATION MODEL OF AUTOMATED NOISE STABILITY ESTIMATION OF SPEECH INTELLIGIBILITY BY LFM-SIGNAL METHOD
V. ZHELEZNYAK, K. RAKHANAU

We consider a simulation model for utomated assessment of a robust method of measuring speech intelligibility LFM-signal. Substantiates the use of the Wigner transform as a frequency-time distribution. A mathematical model of measuring the signal/noise ratio of the Wigner distribution. A method for analysis of variance, which allows you to set the sensitivity of themetrology gauge to determine the range of sensitivity depending on the controlled parameters and influencing factors. With the help of regression analysis are designed analytical presentation of the sensitivity of measuring the signal/noise. Based on this mathematical model of the meter and the back of the sensitivity function created a robust simulation model for automated assessment of speech intelligibility using LFM-signal. The resulting simulation model showed that the use of chirp signals in the system of measuring the automated control of security leakage channels of voice information, the method takes advantage of the harmonic signal and eliminatesthe lack of it caused a significant ripple channel leakage.

THE ANALYSIS OF VISION-BASED METHODS FOR EARLY FIRE DETECTION
N. BROVKO, R. BOGUSH

Reliable and early fire detection on open spaces, in buildings, in territories of the industrial enterprises is important making any system of fire safety. A perspective direction of performance increase for such systems is use of video fire detectors. In the given work the general algorithms for automatic flame and smoke detection on video sequences are considered. The comparative analysis of applied approaches for color segmentation, detection of moving areas, the analysis of spatial changes in lightness and flicker at the edges of flame and smoke is presented; their advantages and lacks are noted. Directions of algorithm development for video fire detection are specified.

ALGORITHMIC FUNDAMENTALS OF COMPUTER MODEL FOR FOREST FIRES PREDICTION
D. BAROVIK, V. TARANCHUK

Forest fires prediction computer model adapted to the conditions, dominant forest species, characteristic scales, topography and climate of Belarus is presented. The basics of input data tables pre-processing, classification, calculation and formation, as well as algorithms implemented in the software complex are described. The proposed two-dimensional model is a generalization of widely used semiempirical one-dimensional Rothermel’s model.

ON X-PERMUTABLE THE MAXIMAL SUBGROUPS OF SYLOW’S AND HALL’S SUBGROUPS WITH THE ALLOCATED SUBGROUPS
А. SHNYPARKOV

Influence X-permutable of the maximal subgroups of Sylow’s and Hall’s subgroups with the allocated subgroups on a structure of finite group is investigated. New conditions of supersolvability and partial super-solvability of a finite group with ?G-supplemented subgroups are received. For the first time H-permutable subgroups have been considered in works [1; 2] where authors receive a number of interesting properties. In particular, it has been noticed that many classes of finite groups can be described in terms of X permutable subgroups.

SUBSTANTIATION OF THE SAMPLE SIZE AND THE CRITICAL VALUE OF THE CRITERION OF THE CONSENT OF THE WHEN TESTING STATISTICAL HYPOTHESES
S. EKHILEVSKIY, O. GOLUBEVA, N. GURIEVA

In work it is shown that the true independent variable defining volume of sample, a kind of criterion of the consent and its critical value, is comprehensible risk. Also there are no the reasons, allowing to minimize risk if for an alternative hypothesis the law of distribution of criterion isn’t concretized. At such approach the significance value and capacity of criterion arise at a stage of intermediate calculations as a by-product and as independent concepts aren't required for the decision of a problem on check of statistical hypotheses.

THE STRUCTURE OF A SCHMIDT GROUPS IN WHICH WITH FIRST, SECOND AND FOURTH MAXIMAL SUBGROUPS ARE GENERALIZED PERMUTABLE
N. HUTSKO, Yu. LUCENKO

Let G be a finite group. A subgroup H of group G is said to be Х-permutable in G (generalized permutation) where X is a non-empty subset of G if for any subgroup T of G there exists an element x of X such that HTx = TxH. The structure of a finite group is closely linked to conditions imposed on the maximal subgroups, Sylow subgroups of the group or the Sylow subgroups of some selected subgroups of the group. In particular, we describe the structure of a Schmidt groups in which every maximal subgroup (generalized) permutes with every 4-maximal subgroup, or every 2-maximal subgroup (generalized) permutes with every 4-maximal subgroup.

SILICON YARDENING AT THE SiO2/Si INTERFACE
D. BRINKEVICH, V. PETROV, V. PROSOLOVICH, N. VABISHCHEVICH, S. VABISHCHEVICH, A. PETLITSKIY

The oxide influence on strength characteristics monocrystal silicon surface layers was investigated by the microindentation method. It was experimentally shown, that the border of SiO2/Si has strengthened layer with microhardness 20 – 35 GPa which in two – three times surpasses size of microhardness, characteristic for volume of a monocrystal. Thickness of this layer was 0,2 – 0,4 microns. Thickness and microhardness of the specified layer depend on conditions of oxide cultivation. Formation of this layer is caused, most likely, by interstitial atoms of silicon formed at border of SiO2/Si at silicon oxidation.

USE DISLOCATION MESOSCOPIC MODELS FOR CALCULATION OF FIELDS OF DISPLACEMENT AND DEFORMATIONS IN SYSTEM “WEDGE KIND NANOTWIN – A FULL DISLOCATION”
O. OSTRIKOV

Mathematical parities for calculation of the deformed condition at wedge kind nanotwin and a full dislocation being near it are received. With use of the given parities fields of displacement and deformations in system not coherent nanotwin – a full dislocation for the first time are calculated. It is established that at twin borders the full dislocation renders essential influence on a configuration of fields of displacement and deformations only at its finding in immediate proximity to the twin. Change of a direction of a vector of Byurgersa of a full dislocation allows to operate a configuration of fields of displacement and deformations at nanotwin. Results have great value in mechanics area twining real materials when twining it is accompanied by the sliding caused by full dislocations.

REFLECTION OF ELECTROMAGNETIC WAVES OF LAYER ENVIRONMENTS
O. ORLOVA, Y. SHCHADENKOV, V. YANUSHKEVICH

The article presents the results of theoretical analysis of interaction electromagnetic waves with hydrocarbon pool. In the work we have shown the criteria of sounding signals characteristics choice, choice of signals used for searching anisotropic formation: the methods of modeling environment over the hydrocarbon pool technical solution, which allow to reduce the weight and size of the radio-engineering systems of searching hydrocarbon pools. The results for qualitative increase of efficiency of search and allocation hydrocarbon pools on the basis of use of electromagnetic waves are shown. There are recommendations for the development of the methods of searching of hydrocarbon deposits. We have worked the methods of searching hydrocarbon pool, based of using optimal correlation of frequencies of electromagnetic waves. The article presents the results of definition of depth hydrocarbon pools. The results of work can be used in the fields of geophysics, radiolocation, electronics.

THE INFLUENCE OF NUMBER AND STABILITY OF INSTANTANEOUS INDICATION OF VOLTAGE ON THE ERRORS AT THE CALCULATION OF UNIFIED POWER QUALITY INDEX (UPQI)
A. VERSHININ, V. AVDEYKO

The results of the investigation of the error of method at the calculation of Unified Power Quality Index in digital meters of UPQI depending on number and stability of instantaneous indication of voltage on the period of the system frequency under synchronous frequency of discretization. The findings are verified with the help of methods of MathCad package for two modelling functions of system voltage with different initial phases in the form of ideal sinusoid and in the form of harmonics mixture under ideal synchronizing and under synchronizing with an error of discretization frequency. The calculation results of the given errors of method are presented in the form of diagrams of relative calculation errors of vertical voltage values and harmonic amplitudes from the number of indications on the period of system frequency, and from synchronizing accuracy of discretization. Conclusions are drawn and requirements to the choice of discretization frequency as well as to the accuracy of its synchronization to the system frequency are formulated on the basis of these results.