Normative improved artificial fish swarm algorithm (NIAFSA) for global optimization

Abstract

Optimization is an important field of research. Various optimization algorithms have been developed to solve optimization problems. Nevertheless, many have not succeeded to achieve the real global optima. Hence, a research on designing and developing a global search and optimization algorithm is presented in this paper. The aim is to enhance the performance of global and local searching strategy in term of best optimal solution. The fish swarm algorithm with the particle swarm optimization with extended memory (PSOEM-FSA) is hybridized with the normative knowledge to become a normative improved fish swarm algorithm (NIAFSA). The feature of global crossover breeding is installed into the proposed algorithm to achieve relatively consistent results. A random initialization of initial population is introduced to spread out the candidates of artificial fishes (AFs) over the solution space. In addition, parameters such as visual and step are made adaptive along the iteration process to balance the contradiction between global and local search ability. The collected results are analyzed and compared with few existing fish swarm variant algorithms to verify the performance of the proposed algorithm.