Software-Enabled Investigation in Metaheuristic Power Grid Optimization

Actual developments in power grid research, analysis, and operation are dominated clearly by the strong convergence of electrical engineering with information technology. Hence, new control abilities in power grids come up that revolutionize traditional optimization issues, requiring novel solution methods. At the same time, heuristic algorithms have emerged to be highly capable of handling those new optimization problems. In this work, a simulation-based optimization approach is proposed that enables investigation with metaheuristic algorithms for domain experts, where especially the power engineering point of view gets highlighted. HeuristicLab is demonstrated as a framework for optimization, which facilitates usage and development of optimization algorithms in a way that is attractive not only to computer scientists. From a software point of view, architectural aspects are treated that enable the decoupling of optimization algorithms and problems, which is a basic fundament of the framework. Further, interprocess communication is discussed that enables the interaction of optimization algorithms and simulation problems, and a practical showcase demonstrates the framework's application to real-world power grid optimization issues.