Switching reconfiguration of fraud detection within an electrical distribution network

The intelligent grid has been envisioned as a next- generation framework to modernize power grid and improve its efficiency and sustainability. Advanced metering infrastructure (AMI) has become indispensable in a smart grid to support the real-time and reliable information exchange. However, computerizing the metering system also introduces numerous new vectors for energy fraud. Energy fraud is a notorious problem in electric power system, which can cause great economic losses in business. This paper presents a graph-theoretic strategy of reconfiguration switching scheme incorporating microgrids to localize potential fraudulent area. First, a graph transformation from a distribution network to a spanning tree is proposed. A mathematical conversion from the graph to adjacency or incidence matrix is then formed. The switching procedures based on the matrix level are utilized to detect fraud in accordance with anomaly score and likelihood of the fraudulent events.