Background: Need a reliable and low cost way to detect changes in the status of switching devices especially circuit breakers in distribution networks. Technology Description: This invention considers a distribution system with circuit breakers and very few sensors which do not need to be located at the circuit breakers. The number of sensors is assumed to be much smaller than is needed for classic state estimation. Using tools from machine learning the proposed invention is able to detect whether the circuit breakers switched from their nominal setting. The invention also includes an analysis methodology that is able to estimate in advance the confidence level This allows for comparison between different sensor placements. With the proliferation of distributed energy sources as well as electric vehicles distribution networks will need to be set up non-radially for optimal operation. Managing such networks requires online topology estimation which is not available in today’s distribution networks. The decreasing cost of sensors while still not low enough for covering distribution networks at the same level as in transmission networks will enable this estimation using the described invention. Applications: Electric power state estimation for distribution grid management.
1) Any automatic procedure to detect outages will lead to much better quality of service and higher revenues. 2) Calculation method is quick and it leads to efficient performance estimation that can take place in advance.