Population Initialization of Seagull Optimization Algorithm with Pseudo Random Numbers for Continous Optimization

This study is presenting a new nature-inspired algorithm termed as Seagull Optimization (SOA) to decipher the complex computational problems. The major concern of this approach is the attacking and migration attribute of seagull. To focus on the exploitation (local search) and exploration (global search) in the defined region, this study proposed two novel initialization approaches of SOA called SOA with SIMD-oriented Fast Mersenne Twister Generator (SOA-SOFMTG) and SOA with Combined Multiple Recursive Generator (SOA-CMRG). The convergence rate of the presented variants has been examined with simulation results and convergence graph of eight benchmark test functions. The effectiveness of proposed variants is investigated with Kruskal-Wallis statistical tests. The computational results conclude that the proposed variants of SOA is capable to resolve the complex optimization problems as compared to the standard SOA.