Software & Systems

Articles of journal № 3 at 2017 year.

Order result by: Public date | Title | Authors

11. An enterprise information system for estimating labor and software development costs [№3 за 2017 год]
Authors: S.L. Kotov, A.A. Demirsky
Visitors: 4360
This article aims to study the process of software size estimation and its impact on the software development process. Software size has the greatest impact on the complexity of development. A software estimation process, which is in-tegrated with the software development process, can help to establish realistic and credible plans to implement the project requirements and satisfy commitments. It also can support other management activities by providing accurate and timely planning information. The article provides an overview of existing methods for estimating the size of the software created. The most common methods are: functional point analysis (FPA) and constructive cost model (COCOMO, COCOMO II). The article indicates disadvantages of these methods. The paper presents a new approach of software size estimation. The fuzzy model Takagi-Sugeno-Kang is described by fuzzy IF-THEN rules, which represent local input-output relations of a nonlinear system. The main feature of the Takagi-Sugeno-Kang fuzzy model is to express a local dynamics of each fuzzy implication (rule) by a linear system model. The fuzzy model Takagi-Sugeno-Kang is used to estimate the size of software objects. The paper illustrates how the new method can be used in the enterprise information system for estimating labor and software development costs. The result shows that the approach is better than FPA method.

12. A software suite to determine the parameters of three-phase alternating current electric arcs burning in a horizontal surface [№3 за 2017 год]
Authors: I.M. Yachikov, E.M. Kostyleva
Visitors: 8792
The metallurgical industry uses a lot of equipment with electrical arcs as the sources of thermal energy. They burn in the direction of the horizontal heat-absorbing surface. Knowing electromagnetic forces affecting the arcs, their location and behavior is significant when designing new furnaces and automating their operation. The literature review shows that many issues connected with electrical arcs of three-phase alternating current are underserved due to technical complexity, high cost and often impossible on industrial or pilot plants. A possible method of research is the development of appropriate mathematical models, their implementation as applied software and further computer modeling. The paper shows mathematical models and developed software, which includes modules for determining the magnetic field intensity, electromagnetic forces, arc and heat flow forms, near three arcs of three-phase alternating current burning in the direction of a horizontal surface. The result of the created software complex is calculation and representation of vector fields including magnetic field strength and average-integral electromagnetic forces, as well as visualization of the average form of electric arc column axis by the period and the heat flux distribution on a horizontal heat-absorbing surface.

13. A concept of intersubjective response receiving to fuzzy queries to forging computer-aided process planning database [№3 за 2017 год]
Authors: Kanyukov S.I., Konovalov A.V., Muizemnek O.Yu.
Visitors: 7976
A usage experience of Computer-Aided Process Planning (CAPP) software for press and hammer forging is a set of technological parameter values obtained in the technological problem solution process. The values are stored in a rela-tional database. State-of-the-art database managing software generally use “crisp” (as opposing to “fuzzy”) databases and allow only “crisp” queries. We use “crisp database” as a term for a set of records that hold attributes of objects of user interest. The attribute values are represented by text or numeric values that are unambiguously understood by users. Crisp queries suppose enumeration of the attribute values or their boundaries. Therefore, the objects with attributes that partly satisfy the query will be excluded from the query response, so the information will be lost. The fuzzy query theory allows overcoming this information loss problem. The theory of fuzzy query to relational database is based on the fuzzy set theory, which was introduced in 1984 by L. Zadeh and then developed by D. Dubois and H. Prade. The paper considers the developed concept for acquiring intersubjective responses to fuzzy queries to relational database of a forging CAPP. The authors use case study of forging CAPP software database queries to define linguistic variables, terms and their quantitative characterizations. There are membership functions for terms mapped to quantitative characterizations. The paper formulates response construction rules according to relative importance of stated goals. The query requests the technological processes with “good” metal elaboration in final forged part and high utilization of metal from ingot. The proposed concept can be applied in development of mechanism for acquiring responses to fuzzy queries in numerous subject areas.

14. Method of frequency-morphological classification of texts [№3 за 2017 год]
Authors: A.А. Osochkin, Fomin V.V., А.V. Flegontov
Visitors: 10980
Appearing of centralized data storages and information accumulation as structured tables or semistructured texts is a result of growing attention to data analysis techniques. The analysis of similar data in the long term allows obtaining important information, which can become a basis for making right management decisions or predicting further development of events in many fields. One of the important directions of such analysis is automatic classification of collected data in electronic form. Its simplified model is reduced to reading, text processing and assigning a topic to a document from a given list of topics. Foreign papers more and more often are devoted to medical data classification for further disease progression forecast on the basis of statistics or diagnosis based on medical history. The main difficulty in classification are natural language texts. They are difficult to classify due to linguistic features of language and support by a part of classification methods of exclusively numerical data. The paper studies scientific activity in the field of NLP based on the annual publication of scientific papers in this field. It also offers the method of Russian-language texts classification that integrates the algorithms of frequency, morphological and intellectual analysis. The paper presents the results of some experiments on the identification method of classification with high classification accuracy. The classification was carried out according to functional, literary, and authorial styles.

15. A software complex for electronic system thermal design: requirements for architecture and functional possibilities of modeling [№3 за 2017 год]
Authors: Madera A.G., Reshetnikov V.N.
Visitors: 8364
The paper considers fundamental concepts of mathematical and computer modeling, which are the basis of the development and creation of a multifunctional software package for thermal design of complex electronic systems. It shows the basic flaws of foreign thermal design software systems, which are not applicable in practice of designing and creating competitive electronic systems. The article establishes the requirements for a multifunctional software package for adequate design of thermal processes and temperature distributions in real electronic systems under real conditions of operation and functioning. Namely, the software package should provide the ability to model thermal processes in electronic systems that are nonlinear, nonstationary, three-dimensional, interval stochastic, as well as take into account the influence of thermal feedback, design and installation of elements in electronic systems, effects of destabilizing mechanical, climatic and radiation factors. The architecture of the multifunctional software package should contain a highly efficient mathematical computing core, a de-veloped service user interface that meets modern requirements for complex software packages and systems. The service shell, in turn, should provide a visual, intuitive and easy-to-understand form of setting initial data in the form of color images of temperature distributions and other thermal characteristics at different hierarchical levels. The software package is implemented in Microsoft Visual Studio that provides an object-oriented approach in the high-level language C# powered by Microsoft Windows operating system family. Modular architecture allows expanding the possibilities and upgrading the software complex effectively.

16. Potential distribution simulation for a dual-gate field silicon on insulator nanotransistor with an asymmetric gate [№3 за 2017 год]
Author: Masalsky N.V.
Visitors: 7200
The paper considers a 2D analytical model of potential distribution and a following threshold voltage model for a thin-film dual-gate field nanotransistor with "silicon-on-insulator" structure with a front gate from two sequentially connected materials with different work function. It also investigates potential behavior in transistor structures in the field of gate lengths less than 50 nanometers, its dependence on drain offsets for different configurations of a frontal gate. The received results show that using two materials with different work function in a frontal gate leads to effective suppression of short-channel effects due to a step function in a potential profile on the boundary of two materials. The shift of a surface line of a surface potential minimum is insignificant with increasing drain biases. The peak electric field at a drain considerably decreases in comparison with similar transistor structure but with a uniform frontal gate. At the same time, it is predicted to be a higher active interelectrode conductance and reduction of subthreshold leak in comparison with classical two-gate field nanotransistors with structure "silicon-on-insulator" in the gate lengths less than 50 nanometers. The distinctive properties of the researched structures are essential lowering of an electrical field peak on boundary of a working area and a drain, reduction of parasitic effects of threshold voltage. The simulation results are in good agreement with experiment data received by means of commercially available software package ATLAS intended for simulation of complex transistor structures. Thus, using two materials with different work function in a front gate of dual-gate field nanotransistors with the structure "silicon-on-insulator" improves their key electrophysical characteristics in comparison with dual-gate field transistors with a uniform front gate and volume analogues.

17. Review of simulation modeling tools for robot groups with social organization elements [№3 за 2017 год]
Authors: M.A. Rovbo, E.E. Ovsyannikova , A.A. Chumachenko
Visitors: 9069
The paper discusses mechanisms that are important for a proper simulation of artificial agents’ group, especially mobile robots, which are organized into a system with the elements of a social structure. The authors provide an overview of simulation tools based on these requirements to determine advantages and disadvantages of existing systems for modeling, simulating and researching such systems. They identify key instruments for a simulation of artificial groups with social structures. The review focuses on simulation tools for biological systems (in particular, ants since they are a good example of a social group), agent-based modeling and robot-specialized simulation tools. The paper emphasizes the level of support of various social mechanisms in simulation tools, as well as their capability to model a large number of agents (in particular, cluster computing support) and the ability to simulate user-provided models and implementations of different aspects of a social group. It was found out that, although there are many general modeling and simulation tools, there is very little support and ready-to-use implementations of even widely used mechanisms, such as pheromones, which are provided by systems to build and research a group of robotic agents with a social structure. On the other hand, the tools that aim at simulating systems with social behavior elements, such as ant modeling systems, are not very scalable or limited in aspects they can simulate. There are no mechanisms to expand them for further research. Based on the review, the paper has detected a lack of a comprehensive simulation system that specifically aims at the research of artificial agent groups with social structure elements. It has also formulated a set of requirements for such a system.

18. Optimization of the problem of conflict detection with dangerous aircraft movement areas to execute on Intel Xeon Phi [№3 за 2017 год]
Author: Rybakov, A.A.
Visitors: 4002
An aircraft generates a wake vortex during its flight. This wake might be dangerous for other air traffic participants, especially in zones with a large number of aircrafts. So, the problem of vortex wake modeling and conflicts detection is very important and in demand. The article considers a vortex wake as a set of dangerous zones in separate traffic segments. A dangerous zone of separate traffic segment is a pencil of spheres – a set of spheres with centers on this segment. Next, we consider the direct mathematical method of detecting an intersection of a line and a pencil of spheres. The line is an immediate trajectory of own aircraft, while a pencil of spheres forms the whole dangerous flight zone. Since the number of traffic segments in a dangerous zone is large enough, it is appropriate to use accelerators in such calculations. The paper considers using Intel Xeon Phi coprocessors for this task and applying some manual sourcecode optimization to enforce execution code vectorization. This approach leads to five-time acceleration of calculation for 107 traffic segments in a dangerous zone.

19. Development and investigation of linear regression models of capacitance of magnetic nanofluids microlayers [№3 за 2017 год]
Authors: T.F. Morozova, M.S. Demin, A.S. Morozov
Visitors: 6544
The paper presents the results of the analysis and synthesis of linear regression models, which describe experimental studies of electric capacity of a cell with a magnetic nanofluid (MNF) microlayer, when changing the following input parameters (factors): a polarizing voltage (0...20 V), a dispersed phase volume concentration (2,3...14,4 %), an interelectrode distance (100...200 μm), layer temperature (293...368 K). The article shows two- and three-dimensional MNF microlayer capacitance curves depending on input parameters. Dependence analysis showed high electrical capacity at ~340 K within the studied concentration range. While changing a dispersed phase concentration from 2,3 % to 14,4 % and using a polarizing voltage, there is also characteristic electrical capacitance maximum which becomes more evident at higher temperatures. With increasing dispersed phase concentration, the magnitude of polarizing voltage increases when the electrical capacitance maximum is detected. To get a mathematical model linking the cell capacitance with MNF and its parameters (factors), the authors used IBM SPSS Statistics (the software for statistical analysis). The obtained regression models take into account the main factors and the effects of their interaction. The paper shows the analysis of the obtained models and estimates their errors (approximation errors). While evaluating errors, the authors considered the minimum, maximum and average errors. Based on these data they made the conclusions on the model appropriateness. The best approximation of the experimental data is in the model that takes into account the effects of factor interaction.

20. Multiprocess system debugging problems [№3 за 2017 год]
Authors: Galatenko V.A., Kostyukhin K.A.
Visitors: 7709
The article considers multiprocess complex systems debugging. The development of high-quality complex systems is a long and time-consuming process. It is believed that debugging takes about a half of this time. Switching to multi-core processor architectures has made parallelism a regular task, as well as errors specific for parallel systems. This makes debugging even more important and at the same time more complicated. The main feature of multiprocessing systems is using complex, asynchronous interactions between system components. This feature influences the approaches to debugging, determines the choice of debugging tools and methods. The complexity of debugging is also determined by the quantity and heterogeneity of multi-process system components, some of them might be hardware. A traditional set of debugging tools is the following: an interactive debugger, a tracer, self-monitoring libraries, reverse execution. Their application changes. The first chapter describes the existing debugging tools and techniques. The authors highlight their advantages and disadvantages. The next two chapters describe the existing problems of multiprocess system debugging. The fourth chapter provides the architecture of multiprocess system debugger. In conclusion authors present the results of their work.

◄ ← Preview | 1 | 2 | 3 | Next → ►