BALDWIN EFFECT BASED CHAOTIC PARALLEL GENETIC ALGORITHM WITH VARIABLE-SCALE LEARNING

摘要

Basing on the biological evolution scheme "random evolution plus feedback", Baldwin effect based chaotic parallel genetic algorithm with variable-scale learning, is proposed by introducing the chaotic mapping and posterior Baldwin learning into the standard genetic algorithm.The chaotic mapping maintains a good diversity of population during evolving; and Baldwin learning with variable-scale speeds up the evolution along the right direction.The performance of new GA is demonstrated on a benchmark problem.Experimental results show that the new algorithm is efficient and robust.