Mind evolutionary algorithms based on knowledge

Mind evolutionary algorithms based on knowledge (KMEA) is proposed in this paper. Rough set theory (RST) and granular computing (GrC) are introduced into MEA to mine and discover the knowledge produced in the process of evolution. Firstly it was used to analyze correlation between individual variables and their fitness function. Secondly, eigenvector was defined to judge the characteristic of the problem. And then the knowledge discovered by RST was used to select evolution subspace and to realize knowledge-based evolution. It takes a big step in imitating human thinking, improving the performance of the MEA effectively. Experiment results have shown that the proposed method has higher searching efficiency, faster convergent speed, and good performance for deceptive problem, multi-modal problem and multi-objective problems.