Density as the Segregation Mechanism in Fish School Search for Multimodal Optimization Problems

Methods to deal with Multimodal Optimization Problems (MMOP) can be classified in three main approaches, regarding the number and the type of desired solutions. In general, methods can be applied to find: (1) only one global solution; (2) all global solutions; and (3) all local solutions of a given MMOP. The simultaneous capture of several solutions of MMOPs without parameter adjustment is still an open question in optimization problems. In this article, we discuss a density segregation mechanism for Fish School Search to enable simultaneous capture of multiple optimal solutions of MMOPs with one single parameter. The new proposal is based on vanilla version of Fish School Search (FSS) algorithm, which is inspired on actual fish school behavior. The performance of the new algorithm is evaluated and compared to the performance of other methods such as NichePSO and Glowworm Swarm Optimization (GSO) for seven well-known benchmark functions of two dimensions. According to the obtained results, presented in this article, the new approach outperforms the algorithms NichePSO and GSO for all benchmark functions.