Skip to Main Content
Genetic algorithms (GAs) are powerful optimization techniques. The optimization performance depends highly on the determination of optimized parameter search spaces, which remain unchanged during GA running. Hence, the objective function evolution may decelerate or even stabilize well before attaining the optimal solution. This article proposes an approach of GAs based dynamic search spaces. It focuses on improving the search space boundaries and allowing GAs to discover new search spaces which are not accessible initially. A GA using this approach is developed and validated to the optimization of power system stabilizer parameters within a multimachine system (16-generator and 68-bus). The obtained results are evaluated and compared with those of ordinary GAs and literature. They show significant improvement in terms of optimization performance and convergence rate.