JOURNAL OF EXPERIMENTAL & THEORETICAL ARTIFICIAL INTELLIGENCE, cilt.36, sa.3, ss.389-414, 2024 (SCI-Expanded)
Real world problems in various domains demonstrate different characteristics of changes over time. This is why several researchers have been interested in dynamic optimisation for the last two decades. Since changes occur over time in a dynamic optimisation problem, the goal of a related algorithm becomes tracking the changing optima over time. Evolutionary algorithms and various swarm intelligence techniques have been adapted in the literature to solve dynamic optimisation problems. The Fireworks Algorithm (FWA) is a recently proposed swarm intelligence algorithm for global optimisation of complex static functions that simulates the explosion process of fireworks. Although a set of improvements over the conventional FWA are presented in the literature for the static optimisation problems, the most evident extension is the Enhanced Fireworks Algorithm (EFWA). In this paper, cost effective extensions of the EFWA are proposed for solving dynamic optimisation problems in continuous space. The performance evaluation of our EFWA-based algorithms is validated with the Moving Peaks Benchmark. Empirical studies on different benchmark instances clearly show the applicability of our extensions. Our EFWA-based extensions outperform the related work in terms of both quality of solution and computational cost for a large set of test instances of the benchmark.