Artificial Cell Swarm Optimization

Abstract

AbstractTraditional meta-heuristic optimization algorithms, such as the particle swarm optimization (PSO) algorithm, genetic algorithm (GA), and bat algorithm (BA) played a vital role to provide impressive near to the optimum solutions for linear/nonlinear complex problems in numerous applications. Nevertheless, in some case, such algorithms may suffer from becoming trapped in local optima with long computational time for convergence. Thus, in order to enhance a broader view over the optimization domain, still further refined studies are carried out to develop these algorithms and to explore new ones based on the inspiration from nature. Thus, a novel meta-heuristic optimization algorithm has been proposed in the present work by employing the concept of artificial cells, which are inspired by biological living cells. An efficient application of artificial cell division (ACD) algorithm has been employed to traverse the search space while decreasing the search time. The inherent properties of ACD algorithm prevent it from premature convergence to local optima. The current work designed a novel artificial cell swarm optimization (ACSO) algorithm. The results compared the proposed algorithm performance to GA, PSO, and the bat algorithm by using seven known benchmark functions. The results established that the performance of proposed ACSO algorithm in terms of the number of iterations required to reach the expected accuracy outperformed the GA, PSO, and the Bat Algorithms. The ACSO achieved the fastest convergence with the benchmark functions with accuracies range 100 or 99% compared to the other optimization algorithms in the current study.KeywordsArtificial cell divisionGenetic algorithmMeta-heuristicParticle swarm optimization algorithmBat algorithmBenchmark functionsSingle objective optimization