Double-Swarm Improved Moth-Flame Optimization Algorithm with Golden Sine for Global Optimization

Abstract

Moth-flame optimization algorithm (MFO) is a swarm intelligent optimization algorithm, which is proposed by simulating the night navigation mechanism of moths. MFO is widely used in various fields since it has a simple structure and easy implementation. However, in the process of MFO search, the population diversity will deteriorate, and premature convergence will occur. To solve the above problems, a double-swarm improved moth-flame optimization algorithm with golden sine operator(DIMFO) is proposed in this paper. The population is grouped by chaos mechanism, and the information exchange between sub-swarms is improved by dynamic recombination. The spiral search and golden sine search are used for the two moth populations respectively, and the Gaussian mutation is utilized to generate flames, which improves the population diversity. In addition, compared with other improved MFO algorithm and advanced optimization algorithms in the classical test set, the results show that DIMFO has advantages in convergence accuracy and speed in most test problems.