Другие журналы

GroshevPublication list: http://elibrary.ru/author_items.asp?authorid=699520
Principles of constructing a scalable Weboriented system for estimating the Paretoapproximation quality in solving the multicriteria optimization problem
Engineering Bulletin # 09, September 2017 УДК: 519.6 Quite often in real situations, the quality of operation of the object or system being examined is not estimated by a single criterion or quality measure, but by a set of such criteria that are equally significant. Tasks of this type relate to, and are an important stage in the design and evaluation of the design object. There are a large number of methods that allow to solve the multicriterion optimization (MCE) problem. A relatively new and progressive class of MCE methods are methods that presuppose the preliminary construction of some finitedimensional approximation of a set of solutions called Paretoapproximation. At present, there are quite a few Paretoapproximation methods, which raises the question of a comparative evaluation of the effectiveness of these methods. As shown in the previous works of the authors, many software systems implementing these algorithms do not provide fast access and do not have an easytouse interface for calculations and analysis. In this paper, the development of a serviceoriented system for the performance and evaluation of the Paretoapproximation quality is considered. The main purpose of this system is to provide fast and convenient access to Paretoapproximation solution and evaluation algorithms, without attachment to a particular operating system or application.
WEBoriented system of Paretoapproximation quality estimation in problems of multipurpose optimization
Engineering Bulletin # 08, August 2017 УДК: 519.6 When solving practical, primarily, technical optimization tasks, the tasks of multiobjective optimization (MCHOtasks) often arise, when it is desirable to exaggerate not one but several objective functions. It is well known that since, as a rule, it is impossible to achieve the extreme value of all objective functions simultaneously, the solution of such a problem is a certain compromise belonging to the Pareto set of the problem, that is, to a set of solutions that are not simultaneously improved for all objective functions. By the solution of the MCEproblem, we call the Pareto set of this problem, assuming that, after constructing some finitedimensional approximation of the given set (Papproximation), the person making the decision (LPR), using informal or formalized methods, selects the unique solution from it. WEBoriented software system PARETOQ is designed for remote evaluation of Paretoapproximation quality in multipurpose optimization problems. The system is designed to solve the problem of consolidation of Paretoapproximation algorithms and algorithms for estimating the quality of approximation. We represent the software used by the system, the architecture of the system, some results of computational experiments.
Combined Pareto Front Visualization Method in MultiCriteria Optimization Based on Hyperspace Diagonal Counting
Engineering Education # 08, August 2016 DOI: 10.7463/0816.0844030 pp. 150164
Using the Andrews Plotss to Visualize Multidimensional Data in Multicriteria Optimization
Engineering Education # 12, December 2015 DOI: 10.7463/1215.0825627 pp. 197214
The PARETO RATING Software System for the Paretoapproximation Quality Assessment in Multicriteria Optimization Problem
Engineering Education # 07, July 2014 DOI: 10.7463/0714.0720253 pp. 193214
Estimate the ParetoApproximation Quality in the Problem of MultiCriteria Optimization. A Review of Programme Systems.
Engineering Education # 04, April 2014 DOI: 10.7463/0414.0709198 The paper makes a statement of the multicriteria optimization problem, gives a brief review of techniques to solve it using its Paretoapproximation, and presents mainly known approximation quality indicators. Principally, the paper gives a review of known programme systems orientated to the Paretoapproximation problem solution, which implement various algorithms to estimate the quality of this approximation.



