Improved Seagull Optimization Algorithm Incorporating Golden Sine and Tent Chaotic Perturbations

Abstract

Aiming at the defects of slow convergence and easy to fall into local optimum of the seagull optimization algorithm, this paper proposes an improved seagull optimization algorithm incorporating golden sine and chaotic perturbation of tent mapping. This algorithm enhances the global search ability through tent chaotic disturbance and Levy flight, accelerates the convergence through golden sine to improve the local search ability. In this paper, the original fixed convergence factor is transformed into a nonlinear decreasing convergence factor to improve the optimization efficiency. The performance is tested on the benchmark functions, and it is used to solving the multiprocessor task scheduling problem. Compared with other algorithms, experiments show that TGSOA has significant improvement over other algorithms in terms of convergence speed and robustness.