A novel swarm intelligence algorithm inspired by the grazing of sheep
Autor: | Vahid Majidnezhad, Mahdi Esmailnia Kivi |
---|---|
Rok vydání: | 2021 |
Předmět: |
Mathematical problem
Optimization problem Fitness function General Computer Science Computer science Computational intelligence 02 engineering and technology 01 natural sciences Swarm intelligence Domain (software engineering) 010101 applied mathematics Algorithmic efficiency 0202 electrical engineering electronic engineering information engineering 020201 artificial intelligence & image processing 0101 mathematics Engineering design process Algorithm |
Zdroj: | Journal of Ambient Intelligence and Humanized Computing. 13:1201-1213 |
ISSN: | 1868-5145 1868-5137 |
DOI: | 10.1007/s12652-020-02809-y |
Popis: | Nowadays, efficient solve of optimization problems is a vital challenge, and computational cost in complicated optimization problems is a motivation to use meta-heuristic algorithms. Nature-inspired algorithms have had good results in these problems. In this paper, a novel nature-inspired meta-heuristic algorithm, namely Sheep Flock Optimization Algorithm (SFOA) is proposed, that mimics shepherd and sheep behaviors in the pasture. The move section of SFOA is consist of three move type (1) shepherd’s guidance, (2) sheep’s interest in previous best experience, (3) sheep’s interest in approaching to other sheep. The grazing section is repeated periodically after several Iterations of the move section. A sheep is a solution, and pasture is the problem’s domain, food measure in each point is the fitness function of algorithm, and target is access to leading food sources. Algorithm efficiency is evaluated by common 23 optimization mathematical problems, including unimodal and multimodal cases and two engineering design problems. The experimental results have proven that SFOA is significantly superior compared to the state-of-the-art meta-heuristic algorithms such as SSA, SBO, GWO, as well as conventional meta-heuristic methods like GA, PSO, ICA. |
Databáze: | OpenAIRE |
Externí odkaz: |