×

You are using an outdated browser Internet Explorer. It does not support some functions of the site.

Recommend that you install one of the following browsers: Firefox, Opera or Chrome.

Contacts:

+7 961 270-60-01
ivdon3@bk.ru

  • Features of automation of the flame detection process by video sequence in the oil and gas industry

    In this paper we propose an algorithm for recognition of the flame on the video sequence. The subject area is the oil and gas industry and its features in the framework of the problem. The current technical means for fire detection are described, the restrictions in their application are characterized. Various approaches to flame recognition on video sequences are mentioned, as well as the need for their joint use to create a reliable model for recognition of the flame on he pictures of video stream.

    Keywords: artificial intelligence, fire safety, automation, color mask, Haar cascade, morphological erosion, oil and gas industry, flame recognition, fire automation, video sequence, each-frame analysis

  • Evolutionary algorithm of search of a set of alternative routes in the conditions of possible influences

    It is shown that the problem of simultaneous processing of dynamic information arrays of different degrees of structure and fuzziness is currently relevant. One of the prototypes of mathematical models containing such information structures is the problem of practical distribution of resources in the conditions of possible, difficult to formalize effects. In this problem there are two factor for the rational allocation of resources: network bandwidth to operating conditions and the preference of the routes of transmission resources on the network in terms of the alleged destructive effects. The high degree of uncertainty inherent in the process reduces the feasibility of resource-intensive distribution algorithms. At the same time, it is necessary to obtain a variety of alternative solutions with diversity in terms of resistance to possible impacts. Since, if all routes pass through one transit node, all of them will be equally exposed to the threats of impact inherent in this node, and when it fails, there will be no alternative routes, which will require re-search of routes for the transfer of resources. Fast heuristics, based, for example, on greedy approaches, can not provide the proper diversity, therefore, even with clear formulations of optimization problems, fall into local Optima. For this reason, it is advisable to Supplement the initial solution formation procedure with borrowed solutions from the previously considered problems. In order to improve the solutions obtained at the stage of formation of the starting population, and to ensure the diversity of the descendants of these solutions, describing the routes of resource transfer, an evolutionary algorithm for finding the set of the shortest time routes of resource transfer. The peculiarity of the process of solving the problem proposed algorithm is to maintain the diversity of the population of solutions to possible threats.

    Keywords: intelligent algorithm, distribution, fuzzy space, adaptation, transport networks

  • Brief overview and software implementation of selected methods for deconvolution of images

    In this article, we briefly reviewed the problem of image quality loss. Methods for restoring defocused images are considered and analyzed. Describes lubrication functions and ways of defocusing the image, as well as a mechanism for eliminating the three main types of image blurring. A number of experiments were conducted on the defocused images. An algorithm for deconvolving an image using a Wiener filter and using the Tikhonov regularization method is disassembled. The analysis of the efficiency of the Wiener filter and Tikhonov regularization for blurred images is performed. The comparative analysis was carried out using the developed software for the restoration of defocused images in the Microsoft Visual Studio 2012 environment. For the Fourier transform, the library was used - aForge. A certain dependence of the execution time of the algorithm on the size of the image to be reconstructed. It is established that on the tested problems - the time complexity of the Wiener filter is 1.1 times less than the time complexity of Tikhonov regularization.

    Keywords: Wiener filter, Wiener deconvolution, Tikhonov regularization, image reconstruction, blurred images, motion blur, Gaussian blur

  • Agent reconfigurable fuzzy heterogeneous in space solutions

    The aim of this work is to develop a model of the agent changing its own structure, and movable in the fuzzy heterogeneous search space. The need of creating these search tools solutions is caused not only by lack of data on the solved optimization tasks, but also complex computational structures used in modern information systems. The agent is represented as a point in a fuzzy heterogeneous search space, the scale of each axis of the space built on the basis of the corresponding fuzzy sets. The movement of the agent along each axis is based on the permissible operations on the elements of the corresponding fuzzy sets. An example of agent movement along the axis specified on the basis of S-fuzzy sets. To determine the stopping agent developed automatic adaptation. It is shown that changing the position of the agent in the search space not always leads to the change of uncertainty inherent in the decision, described by the agent. The novelty of the approach is to develop adaptive algorithms move the agent in a heterogeneous search space. On the basis of machine adaptation is shown the scheme of movement of the agent in fuzzy search space along the axis specified with S - fuzzy sets.

    Keywords: intelligent agent, heterogeneous structures, fuzzy space, adaptation, S-fuzzy set, automatic adaptation

  • Overview of content clustering algorithms in image retrieval tasks

    Content Based Image Retrieval is a one of the most popular search methods used in large-scale databases. Comparison the query image with each image of large scale database leads to lowering the systems performance. In this article are considered different clustering algorithms. They may be used for optimizing image searching access period.

    Keywords: CBIR, clustering, K-means, fuzzy C-means (FCM), Possibilistic C-means (PCM), bio inspired algorithm

  • Adaptive bionic algorithm for solving the problem of data flow minimum cost

    Presents an adaptive algorithm for solving the data flow of minimum cost in a static and a dynamic formulation. In the dynamic formulation of the problem change the matrix describing the network. An important component of the algorithm is to use the ideas of co-evolution, the choice of models of evolution (micro-, macro-, meta-evolution), adaptation to the external environment, hierarchical management of genetic and evolutionary search, local search solutions and the use of all modified by genetic operators based on greedy strategies and search methods. Given the example of the recommended data flow based on a known formula the definition of fuzzy proximity µx(b) variable b to the specified value. The adjustment of the process data under the recommended settings implemented with the help of machines adaptation. A distinctive feature of the algorithm is the use of machines adapted for determining the need for and the method of modifying intermediate solutions, as well as for a decision about modifying the previously obtained solutions.

    Keywords: data flow, adaptation, evolution, optimization, evolutionary search

  • One method of constructing database queries based on fuzzy logic

    It is shown that the current direction of the development of electronic document management is the transition to the information systems that support the dialog interaction with the user in natural language. This area is challenging because of exponential growth of data processed by modern information systems. Described the problem of converting verbal criteria describing properties of the required data, SQL queries, implementation of which will allow to obtain the required data. As a practical example, the selected task is the classification of employees in three clearly defined categories: ""small experience"", ""average experience"", ""great experience"". Before you create a database assumptions were adopted: the experience of employees is taken into account from 0-25 years. The time period to 7 years is an indicator of employees with low seniority. As close to a group of employees with secondary seniority are those who have work experience of 8 to 13 years; a great experience – more than 15 years. Create table ""Employees"", consisting of the following fields: id, name, Sex, Phone, notes, date of hiring, date of dismissal, work Experience in similar position, Experience just. The work deals with the queries to retrieve data based on the SELECT statement of the SQL language. Mathematically formalizing verbal filter criteria data was formalized with trapezoidal membership functions. Describes the implementation of the conversion process trapezoidal membership functions in SQL queries against a relational database. Implemented conversion mathematically consistent as the theory of fuzzy sets and the concept of relational databases (i.e. SQL standard). Compliance transformation of fuzzy set theory in the future will allow the use of operators for manipulating sets (Union, intersection, etc.), and compliance with SQL to practically implement the proposed approach by means of a broad class of relational database management systems.

    Keywords: fuzzy logic, membership function, fuzzy set term, linguistic variable, database, fuzzy queries, the index of conformity,a-slice

  • A study of the options adaptive analysis of solutions to optimization problems on the basis of logics Reichenbach and Lukasiewicz

    One of the components of the optimization problem is a set of constraints describing the basic requirements for solutions. So as to find a solution that fully satisfies all the wishes of the experts, it is not always possible, the search area extends through analysis of semi-feasible solutions. The adjustment of the system constraints may contribute to changes in the structure and appropriateness of the decisions. The paper shows the ways to form a generalized membership functions of vague constraints of optimization problems based on the logics of Reichenbach and Lukasiewicz. It is known that for the same design procedures in some cases it is necessary to obtain accurate solutions, while others just get approximate solutions. Therefore, we analyzed the features of membership functions obtained using these logics. It is shown that by implication deny rules to allow logic-based Reichenbach membership function takes values equal to one if the value of the function allow rule is one, or if the value of the function deny rule is zero. By implication deny rules in allowing based on Lukasiewicz logic membership function takes values equal to one if the value of the function allow rule is greater than the value of the function deny rules. Therefore, it can be argued that when designing systems with increased reliability (precision) is more expedient to use the function implications for Reichenbach deny rules in a permissive compared with the same implication by Lukasiewicz. The implication deny rules to allow for the Lukasiewicz appropriate to use when designing subsystems that perform secondary functions that are not systemically forming, etc.

    Keywords: adaptation, fuzzy system, the implication, intellectual method, membership function, optimization, logic, Reichenbach, Lukasiewicz logic