Genetic Algorithm for Optimal Meter Placement and Selection in Distribution Grid State Estimation

State estimation in distribution grids is crucial for various smart grid applications that rely on the current grid state. Due to the high number of possible measurement positions in combination with different measured quantities, such as voltage, current or power, in distribution grids a comprehensive coverage is costly. However, since the estimation quality is related to the number of measured variables, an optimal set of measurement devices has to be determined that also considers the costs. Therefore, in this paper a method for optimal placement of measurements based on a genetic algorithm is presented. The algorithm is tested in exemplary low voltage grids and the results are analysed using the Pareto frontier to illustrate the trade-off between estimation accuracy and cost of measurements.