A Real Coded Genetic Algorithm with an Explorer and an Exploiter Populations

We introduce the concept of a bi-population scheme for real-coded GAs (bGAs) consisting of an explorer sub-GA and an exploiter sub-GA. The explorer sub-GA mainly performs global exploration of the search space, and incorporates a restart mechanism to help avoid being trapped at local optima. The exploiter sub-GA performs exploitation of fit local areas of the search space around the neighborhood of the best-so-far solution. Two types of bGA are presented, aimed at addressing different classes of problems. We also explore a method for adaptive load balancing between the two sub-GAs within a bGA, which uses knowledge of the number of restarts occurring in the explorer sub-GA. The proposed technique exhibits performance significantly superior to standard GAs on two complex highly multimodal problems.