An improved quantum-behaved gravitional search algorithm for high-dimensional multi-modal optimization

摘要

A quantum-behaved gravitional search algorithm based on levy flight (LQ-GSA) is proposed on the basis of analyzing the mechanism of quantum-behaved gravitional search algorithm (QGSA), aiming at solving high dimensional multimodal optimization problems. Firstly, an adaptive dynamic adjustment strategy is proposed for the unique parameter of QGSA—contraction and expansion coefficient (CE), so as to maintain the diversity of population evolution. Secondly, levy flight strategy is introduced in the process of particle location update to expand the search range of particles and enhance the ability of particles to jump out of local optimal. Finally, through the optimization experiment results of six standard test functions in different dimensions, it is shown that LQ-GSA is significantly better than other comparison algorithms in terms of convergence accuracy, convergence speed and stability. With the increase of dimension, the advantages become more prominent, the algorithm shows better performance in solving multi-dimensional and multi-modal optimization problems.