An Evolutionary Framework for Large-Scale Constrained Optimization

In recent decades, large-scale optimization has received significant research attention; however, most of these studies have not considered problems with functional constraints. The introduction of constraints significantly amplifies the difficulty of solving optimization problems. Given the preva-lence of high-dimensional constrained optimization problems in real-world applications, a critical need has emerged for an in-depth exploration of this research domain. This paper presents a novel framework that can tackle complex, large-scale constrained optimisation problems. The framework incorpo-rates a decomposition method that leverages interactions among decision variables, employing a contribution-based strategy to prioritize subproblems that have more substantial influence on enhancing solution quality. Furthermore, the framework integrates constraint consensus to mitigate constraint violations throughout the search process. The proposed algorithm is evaluated on a test suite of constrained overlapping problems, revealing its superior performance when compared to other state-of-the-art algorithms.