Adaptive crossover and mutation in genetic algorithms based on clustering technique

摘要: Instead of having fixed px and pm, this paper presents the use fuzzy logic to adaptively tune pm for optimization power electronic circuits throughout process. By applying K-means algorithm, distribution population in search space is clustered each training generation. Inferences are performed by a fuzzy-based system that fuzzifies relative sizes clusters containing best worst chromosomes. The proposed adaptation method applied optimize buck regulator requires satisfying some static dynamic requirements. optimized circuit component values, regulator's performance, convergence rate favorably compared with GA's using p.