| Summary: | In dynamic constrained optimization, changes may occur in either the objective function or constraint functions, or both. However, although research on dynamic optimization has been growing significantly, it is centered mainly around unconstrained problems. On the other hand, research on dynamic constrained problems has considered simple extensions of those conducted on unconstrained ones. These approaches are usually computationally expensive and exhibit slow convergence in changed environments. To develop an effective algorithm for dynamic constrained problems with changes occurring in only the constraint's space, in this research, we propose a sensitive constraint detection mechanism that provides valuable information for determining the movements of solutions in changing environments. As, by incorporating it with a search process, the convergence of an algorithm can be accelerated, it is adopted with a multi-operator evolutionary type of algorithm which is then tested on a set of constrained dynamic problems. The experimental results clearly demonstrate the benefits of this proposed approach.
|
|---|