Solver schemes of partial differential equations can be adapted through parameter selection. However, this selection is often times done in a manner of try and error. We present a new approach to automate and optimize the parameter selection that gives us the opportunity to reuse data as requirements and optimization goals change.