In standard cultural algorithm

the influence function guides the evolution only by a single layer belief space

which may invalidate the structure of cultural algorithm and lead to poor global optimization and instability.Therefore

a cultural algorithm based on multi-layer belief space is proposed.The algorithm increases the validity of knowledge and at the same time provides the optimal mode for the evolution of main population space by managing multilayer belief spaces hierarchically.And adjust the fusion mechanism of the multilayer belief spaces adaptively according to the population dispersion so that it can keep the population diversity in the early evolution and accelerate population convergence in the late evolution.Aiming at benchmark functions

comparing with some other similar algorithms

the proposed algorithm shows better optimization performance in the precision of the solution

stability and global convergence.