Selective Pressure Strategy in Multilevel Cooperative Coevolution for Solving LSGO Problems

Abstract

Continuous large-scale global optimization (LSGO) is a challenging task for a wide range of state-of-the art metaheuristic algorithms. This is due to the curse of dimensionality because the size of the search space grows exponentially with the increasing the number of variables. Thus, metaheuristics lose efficiency in solving LSGO problems. For increasing the performance of metaheuristic algorithms in solving LSGO problems, cooperative coevolution (CC) is used. CC-based metaheuristics have two main control parameters, which are the population size and the type of variable grouping. In this paper, a novel selfadaptive multilevel cooperative coevolution algorithm is proposed. The subcomponent optimizer of the proposed CC-based algorithm is SHADE. The proposed algorithm self-adapts the number of subcomponents and the population size during the optimization process. The complete title of the proposed algorithm is CC-SHADE-ML. We have evaluated the performance of the proposed algorithm on fifteen benchmark problems chosen from the LSGO CEC'2013 benchmark set. The performance of the CC-SHADE-ML algorithm has been evaluated using well-known mutation strategies and selection operators. We can conclude that mutation and selection operators make a significant impact in the performance of DE-based metaheuristics. All numerical experiments are proven statistically.