A Review of Recent Advancements of Firefly Algorithm; A Modern Nature Inspired Algorithm

Abstract

Nature inspired algorithms are modern AI related algorithms which embrace natural optimization techniques to build optimization algorithms. This new field has attracted many researches and one can now see a rapid development of the algorithms as well as frequent application of the algorithms on complicated problems. Firefly Algorithm is one such algorithm that mimics the flashing behaviour of fireflies. The algorithm which was initially proposed for continuous domain optimization problems has been adopted for many types of real world continuous optimization problems. Further, many research has been done on adjusting the algorithm for various discrete optimization problems. This paper aims to provide a review of the recent advancements in the firefly algorithm and the type of problems addressed by the research on continuous and discrete domains. Applications of firefly algorithm were selected from both domains and several important factors like the problem of interest; comparisons with other algorithms and new modifications to the original algorithm were taken into account when doing the study. The most important factor is how the researchers have used exploration and exploitation properties in Firefly algorithm for the problems they have solved. A brief analysis has also been carried out to compare the implementation of the same problem by different nature inspired algorithms, including the firefly algorithm. The results of the review reveal that in many applications, firefly algorithm’s performance is remarkable compared to other nature inspired algorithms like Genetic algorithms. Firefly algorithm’s behaviour is similar to genetic algorithms but the algorithms like Bat, cuckoo search are more likely to behave as particle swarm optimization, where always the globally best solution is concerned. The study concludes that firefly algorithm’s performance over many applications is admirable and is worth modifying to solve many real world complex optimization problems.