Topology identification in distribution network with limited measurements

We consider here the problem of detecting changes in the status of switching devices, circuit breakers in particular, in distribution networks. The lack of measurements in distribution networks compared to transmission networks is the main challenge of this problem. Using expected values of power consumption, and their variance, we are able to quickly calculate the confidence level of identifying the correct topology, or the current status of switching devices, using any given configuration of real time measurements. This allows to compare between different configurations in order to select the optimal one. The main approach we propose relies on approximating the measurements as normal distributed random variables, and applying the maximum likelihood principle. We also discuss an alternative based on support vectors. Results are demonstrated using the IEEE 123 buses distribution test case.