Self-Adaptive Gravitational Search Algorithm With a Modified Chaotic Local Search

Abstract

The gravitational search algorithm (GSA) has been proved to yield good performance in solving various optimization problems. However, it is inevitable to suffer from slow exploitation when solving complex problems. In this paper, a thorough empirical analysis of the GSA is performed, which elaborates the role of the gravitational parameter $G$ in the optimization process of the GSA. The convergence speed and solution quality are found to be highly sensitive to the value of $G$ . A self-adaptive mechanism is proposed to adjust the value of $G$ automatically, aiming to maintain the balance of exploration and exploitation. To further improve the convergence speed of GSA, we also modify the classic chaotic local search and insert it into the optimization process of the GSA. Through these two techniques, the main weakness of GSA has been overcome effectively, and the obtained results of 23 benchmark functions confirm the excellent performance of the proposed method.