SCIENCE & EDUCATION

 The paper introduces a concept “expanded semantic network of a tutoring system”. Also the paper discusses a number of complexity measures of this network - complexity measures for concept, shareable content object, object repository and content package 

The paper discusses a class of problems with two-level tree information structure graph with leaves’ computational complexity a priori unknown. Using LSF component of a well-known GRID-system Globus for distributed heterogeneous computing system load balancing while parallelizing problems of the specified class is described.

This paper is devoted to solve multi-objective decision making problems. The approach proposed is based on two-stage procedure, which utilize Tabular Method and well-known Analytic Hierarchy Process to choose the best alternative among the given ones. Software tools based on the proposed approach are developed by authors, therefore the approach becomes usable to utilize in a practice by anyone.

“Experts command” and “virtual particles” algorithms used for artificial neuron network training, as well as their concurrent implementation at cluster architecture with the use MPI functions library are examined in the article. Outcomes of the experiment carried out on the basis of Krasnoyarsk State Technical University cluster are set out.

     Definition method of output impedance admissible values of the independent object power-supply system at early design stages in different making conditions of electric consumption channels is described. Way of lowering of the requirements to the output impedance of the power-supply system is offered.

In this article on basis of fuzzy logic stages of development of a model of a carbonizing column being used in producing soda ash by ammoniac method are considered.

There is made an example of the relation not satisfying the third normal form in the article. Informal and then formal algorithms of functional dependency detection are proposed. The informal algorithm of functional dependency exclusion is proposed. Manipulation sequence with the relation for its cast to the 3rd normal form is demonstrated by way of an example. Application features of legacy database design tool for not great dimension relations cast to the 3rd normal form are considered.

The algorithm of deliverance from complex attributes is proposed in the article. Application features of legacy database design tool for not great dimension relations cast to the 1st normal form are considered. The algorithm of subheadings exclusion inside a table is demonstrated by way of an example. The algorithm of internal subheadings exclusion is proposed. Application features of legacy database design tool for internal subheadings exclusion in not great dimension relations are considered.

There are considered problems of creation of the system of insurance protection of industrial enterprises in the article. The positive and negative aspects of using insurance as a mechanism of risk management are disclosed. The purposes and principles of functioning of the system, features of strategy of the industrial company in the field of maintenance of insurance protection, practical experience of “Lukoil" company are analyzed.

Calculated research of the heat input inopportuneness coefficient under characteristics of allocation and removal of heat in a piston engine are resulted here. The analysis of determining factors of size of the heat input inopportuneness coefficient and its formation in development of a cycle is given in this work. In a basis of research is put the accumulated experience of use of a method of the analysis of the internal-combustion engine indicated efficiency.

In this paper matrix and analytic methods of graph structural models representation are analyzed and systematized. Methods of hierarchic models representation are suggested. Relationship features of model elements are determined and representation methods are characterized.    

The information system «Vocational guidance» is developed for problems of vocational guidance solving.  Diagnostics of professional gift in the information system is made by means of the expert system entering into its structure. The productional model is chosen as a model of knowledge representation. Several experts’ opinions are considered by means of an expert judgement method for reduction of a subjectivity degree. For knowledge uncertainty processing it is offered to use the combined method on basis of three methods: Chris-Naylor, Shortleaf and Buchanan, Dempster-Sheifer. To E.L. Klimov’s classifier the additional group of trades uniting interdisciplinary specialties is added. Entrants monitoring is carried out in the subsystem of the cartographical analysis with use of GIS-technologies.

The problem of complex system model reconstruction on incomplete data, in particular, on the experimentally observed signal is considered. Features of wave representation of registered biosignals are investigated. The main stages of creation of the model equations of the complex system which output signals show non-stationary wave character are analyzed. The linear model reconstruction algorithm for the pulse mechanism on the basis of the wave description of a pulse signal is presented. Examples of the basic function choice are given. Comparison with reconstruction algorithms of nonlinear models of complex systems is carried out.

The author carried out a retrospective analysis of influence of different factors on the value of the temperature recovery factor which affects, in certain way,  gas dynamic temperature lamination. It is shown that the value of the temperature recovery factor could be determined as a square root of the Prandtl number in case of a longitudinal laminar gas flow over simple bodies (plane, cone, and cylinder) within a wide range of parameters. In case of the turbulent longitudinal flow, the temperature recovery coefficient almost doesn’t depend on the Mach and Reynolds numbers and can be determined as a cubic root of the Prandtl number within the range from 0.5 to 2. It is also shown that the value of the temperature recovery factor decreased in case of an airflow over complex bodies in the presence of blowing; but this value could be determined only experimentally.

This works describes the test configuration of gas dynamic energy disjunction device (Leontiev’s tube) working on the natural gas. This device works by means of local disbalance between heat generation and dissipation during the flow of compressible gas. Results of experimental investigation of the specified device and measuring error estimate were presented in this article. Updated method of calculating this class of devices was proposed; it allows one to consider the intensification of heat transfer at the initial section of a supersonic channel along with the influence of the longitudinal pressure gradient and the longitudinal thermal conductivity of this device’s walls. The comparison of the results obtained from numerical experiments with a use of updated method with the results obtained from the same experiments with a use of basic method and experimental data was carried out. It was also shown that results of this research obtained using the updated method describe the growth of natural gas’ enthalpy better than ones obtained with a use of basic method. 

Various problems of prediction in tool production are considered and approaches to their decision are investigated. Features of the accounting of non-stationary production factors and the hidden interrelations are identified. Need of application of artificial neural networks is proved. Algorithms of prediction of equipment loading, need for a material, the forecasting of marriage and energy consumption which are realized by means of a neuronet of direct distribution are offered. Examples of prediction problems in an information and analytical control system of supply and production of the tool are given.