Abstract
Hyper-Heuristics are high-level methodologies developed to select or generate heuristics for solving complex problems. Despite their success, there is a lack of multi-objective hyper-heuristics. In the multi-objective optimization context, MOEA/D decomposes a problem into a number of sub problems handled by individuals in a collaborative manner. Our approach, named MOEA/D-HHSW, expands the MOEA/D framework with a multi-objective selection hyper-heuristic. It uses the proposed adaptive choice function with sliding window to determine which low-level heuristic (differential evolution operators) should be applied by each individual during MOEA/D execution. The proposed approach is tested in three known instances of the multi-objective environmental/economic dispatch problem, formulated as a non-linear constrained optimization problem with competing and non-commensurable objectives. MOEA/D-HHSW outperforms state-of-the-art algorithms reported in the literature for all considered instances.