Experimental study on solution possibilities of multiextremal optimization problems through heuristic methods

Abstract

The work objective is to study a vital task of the multiextremal objects search engine optimization which is much more complicated than monoextremal problems. It is shown that only heuristics is appropriate in achieving this goal. Therefore, three best known and developed search engine optimization techniques are studied: particle swarm method, evolutionary genetic approach, and ant colony algorithm. The analysis is performed in the environment common for all methods of the test research problems of the multiextremal Rastrigin function. It is proved that all these methods are well suited for the multiextremal problem solution. While it is necessary to use proper specific approaches to solving the local extremum detection and identification problem in each of the heuristic algorithms, they all require data clustering. Each method can provide any desired accuracy of the extremum problem solution, and it utilizes an acceptable time resource.

Keywords

• оптимизация
• экстремум
• многоэкстремальность
• поисковая оптимизация
• кластери-зация
• эвристические методы
• эволюционно-генетический подход
• метод роящихся частиц
• муравьиный алгоритм
• optimization
• extremum
• multiextremality
• search engine optimization
• clustering
• heuristic methods
• evolutionary genetic approach
• particle swarm method
• ant colony algorithm