| Autor: |
Gelareh Veisi, Parisa Molavi Damanahi, Seyyed Javad Seyyed Mahdavi Chabok |
| Rok vydání: |
2015 |
| Předmět: |
|
| Zdroj: |
2015 International Congress on Technology, Communication and Knowledge (ICTCK). |
| Popis: |
In most of the problems, finding global and local optima in multi-modal optimization problems are important. Standard Meta-heuristic algorithms are converging to an optimum, while the goal is finding all optima. (DE) is a well-known and powerful optimization. In this paper we proposed a novel method to consider solving the problems of multimodal in high dimensional by very good accuracy. Mechanism of proposed approach is that roaming method is used for making parallel sub-populations. Then several improved proposing strategies and VPSDE method are assigned to each sub-population randomly. Furthermore, we used Hill-Valley method for detecting whether two points are in same species or not. our proposed approach is compared with well-known state-of-the-art niching algorithms and results show that our approach outperforms all of them. |
| Databáze: |
OpenAIRE |
| Externí odkaz: |
|
