Статья посвящена анализу методов распараллеливания метаэвристик с целью выбора наиболее подходящей комбинации «метаэвристика — метод распараллеливания» для решения задач децентрализованного диспетчирования, в частности — задачи формирования конфигурации информационно-управляющей системы. Рассмотрены: имитация отжига, поиск с запретами, генетические алгоритмы и алгоритмы муравьиных колоний. Обоснован выбор метода распараллеливания и семейства алгоритмов, рассмотрены перспективы их применения в системах реального времени.
ABSTRACT
The article is devoted to the analysis of methods of metaheuristics parallelization in order to choose the most relevant combination of metaheuristic and method of parallelizing for decentralized dispatching problem solving and problems of configuration formation of the information-control system in particular. Simulated annealing, tabu search, genetic algorithms and ant colony algorithm are considered. The selection of parallelization method and family of algorithms is proved; prospects of their use within real-time systems are determined.
Candidate of Technical Sciences, junior research scientist, Research Studies Institute of Multiprocessor systems named after the academician A.V. Kalyaev Federal State Autonomous Educational Institution of Higher Professional Education “Southern Federal University”, Taganrog, Russia
Candidate of Technical Sciences, research scientist, Research Studies Institute of Multiprocessor systems named after the academician A.V. Kalyaev Federal State Autonomous Educational Institution of Higher Professional Education “Southern Federal University”, Taganrog, Russia
Programmer, Research Studies Institute of Multiprocessor systems named after the academician A.V. Kalyaev Federal State Autonomous Educational Institution of Higher Professional Education “Southern Federal University”, Taganrog, Russia
ISSN 2311-5122. Article metadata is hosted on the eLIBRARY.RU platform. Publisher — LLC «MCNO» Editor-in-Chief - Marina Yu. Zvezdina.
By continuing to use the site, you agree to the processing of cookies and user data collected, including via Yandex.Metrica, for the purpose of site operation and statistical research. If you do not want your data to be processed, change your browser settings or leave the site.