---

This thesis introduces a comprehensive methodology for the automation of the strategic power system planning process in the medium voltage level. The methodology takes the predicted development of load and distributed generation as well as the age structure of the components into account. Target grid structures are computed with a heuristic search that considers constraints for the grid topology, power flow parameters in normal as well as contingency operation, fault currents and service reliability. The implementation is based on the newly presented open source power systems analysis tool pandapower, which allows grid modelling and analysis with a high degree of automation.
The developed methodology is applied to three real case study grids from different power system operators. The structural optimization leads to a reduction of investment and operational costs within the planning horizon of up to 56% in the target grids compared to the present grid structures. The successful application of the developed method to a diverse set of case studies demonstrates its general applicability in realistic planning problems.

---