Rationalized fruit fly optimization with sine cosine algorithm: A comprehensive analysis. (1st November 2020)

• Record Type: Journal Article
• Authors: Fan, Yi; Wang, Pengjun; Heidari, Ali Asghar; Wang, Mingjing; Zhao, Xuehua; Chen, Huiling; Li, Chengye
• Is Part Of: Expert systems with applications. Volume 157(2020)
• Publication Date: 2020-11-01
• Subjects: Fruit fly optimization algorithm -- Sine cosine algorithm -- Global optimization -- Swarm intelligence -- Kernel extreme learning machine; Expert systems (Computer science) -- Periodicals; Systèmes experts (Informatique) -- Périodiques; Electronic journals; 006.33
• DOI: 10.1016/j.eswa.2020.113486
• Languages: English
• ISSNs: 0957-4174

Abstract

Graphical abstract: Highlights: A sine cosine-based FOA (SCA_FOA) is proposed. A comprehensive set of benchmark functions are involved for comparison. The effectiveness of SCA_FOA is confirmed over the state-of-the-art algorithms. SCA_FOA is also tested over practical problems from IEEE CEC 2011. Abstract: The fruit fly optimization algorithm (FOA) is a well-regarded algorithm for searching the global optimal solution by simulating the foraging behavior of fruit flies. However, when solving high dimensional mathematical and practical application problems, FOA is not competitive in convergence speed, and it may quickly fall into the local optimum. Therefore, in this paper, an enhanced fruit fly optimizer, termed SCA_FOA, is developed by introducing the logic of the sine cosine algorithm (SCA). Specifically, in the process of searching for food utilizing the osphresis organ, the individual fruit fly adopts the way inspired by the SCA to fly outward or inward to find the global optimum. A comprehensive set of 28 benchmark functions were used to measure the exploitation and exploration abilities of the proposed SCA_FOA. The results demonstrate that SCA_FOA is superior to other competitive algorithms. Moreover, 10 practical problems from IEEE CEC 2011, three engineering problems, three shifted and asymmetrical functions, and optimization problems of kernel extreme learning machines (KELM) were also solved, effectively. The results and observations indicate that not only the proposed SCA_FOA can be used for simulated problems as a very efficient method, but also it can be employed for real-world applications.