Enhancing firefly algorithm with sliding window for continuous optimization problems

• Published in: Neural computing & applications Vol. 34; no. 16; pp. 13733 - 13756
• Main Authors: Peng, Hu, Qian, Jiayao, Kong, Fanrong, Fan, Debin, Shao, Peng, Wu, Zhijian
• Format: Journal Article
• Language: English
• Published: London Springer London 01.08.2022; Springer Nature B.V
• Subjects: Algorithms; Archives & records; Artificial Intelligence; Computational Biology/Bioinformatics; Computational Science and Engineering; Computer networks; Computer Science; Data exchange; Data Mining and Knowledge Discovery; Data transmission; Experiments; Heuristic methods; Image Processing and Computer Vision; Microbalances; Network reliability; Optimization; Original Article; Parameter estimation; Probability and Statistics in Computer Science; Sliding; Swarm intelligence; Swarm intelligence algorithm; Firefly algorithm; Attraction model; Sliding window
• ISSN: 0941-0643; 1433-3058
• DOI: 10.1007/s00521-022-07193-6

Firefly algorithm (FA) is a simple and effective swarm intelligence algorithm, which has received wide attention from scholars. In original FA, each firefly must be compared with other fireflies in brightness, but it may not move, which may result in waste of system resources. Therefore, an enhancing firefly algorithm with sliding window (SWFA) is proposed in this paper to address the above problem. SWFA introduces sliding window mechanism to improve the attraction model of the FA, which is a technology used to ensure the reliability of data transmission in computer networks. The sliding window mechanism is essentially an archive mechanism, where the window denotes a form of archive, and sliding is the way the window updates. The update of the population is guided through the method of information exchange among individuals inside and outside the window. SWFA also combines the sliding window mechanism with reverse learning to reduce the number of comparisons and ensure every comparison is effective. Moreover, a novel adaptive step adjustment strategy is designed, which balances exploration and exploitation of FA. In order to verify the effectiveness of SWFA, extensive experiments are conducted on the CEC 2015 and CEC2013 test suite. Additionally, experiments are conducted on parameters estimation of chaotic systems and three practical engineering optimization problems. The results of the experiments show that the proposed algorithm has better performance.