Previous [ 1] [ 2] [ 3] [ 4] [ 5] [ 6] [ 7] [ 8] [ 9] [ 10] [ 11] [ 12] [ 13] [ 14] [ 15] [ 16]

Journal of Information Science and Engineering, Vol. 32 No. 3, pp. 661-676 (May 2016)

Improved Speciation-Based Firefly Algorithm in Dynamic and Uncertain Environments

1Department of Computer and Information Technology Engineering
Islamic Azad University
Qazvin Branch, Qazvin, 34185-1416 Iran
2Department of Computer Engineering and Information Technology
Amirkabir University of Technology
Tehran, 15875-4413 Iran

Many real-world optimization problems are dynamic in nature. The applied algorithms in this environment can pose serious challenges, especially when the search space is multimodal with multiple, time-varying optima. To address these challenges, this paper proposed a speciation-based firefly algorithm to maintain the population diversity in different areas of the landscape. To improve the performance of the algorithm, multiple adaptation techniques have been used such as adapting the number of species, number of fireflies in each specie and number of active fireflies in each specie. A set of experiments are conducted to study the performance of the proposed algorithm on Moving Peaks Benchmark (MPB) which is currently the most well-known benchmark for evaluating algorithm in dynamic environments. The experimental results indicate that the proposed algorithm statistically performs better than several state-of-the-art algorithms in terms of offline-error.

Keywords: firefly algorithm, multi-swarm, dynamic optimization, speciation, dynamic and uncertain environments

Full Text () Retrieve PDF document (201605_09.pdf)

Received December 21, 2014; revised February 18, 2015; accepted March 26, 2015.
Communicated by Tzung-Pei Hong.