Network reduction algorithm for smart grid applications

The increasing number of distributed energy resources on distribution level causes several problems in distribution grids like overloads and voltage range deviations. To resolve these problems and to avoid a necessary cost intensive network expansion, smart grid solutions can be applied. However, the lack of measurement data on the distribution level means that there is no full knowledge of the current grid state. Thus, smart monitoring and control systems need to estimate the current grid state in order to identify appropriate measures for eliminating critical states. In this paper a network reduction algorithm is presented, which aggregates unobserved network areas. Together with measurements of critical nodes a full mathematical model can be derived, which enables to run a power flow calculation for grid state identification. The proposed method has been tested on a small 15 node test system and on a real middle voltage 130 node grid. Simulation results show an adequate accuracy of the introduced approach.