Real time electricity theft detection in microgrids through wireless sensor networks

A microgrid (MG) is potentially vulnerable to electricity theft when it operates in three different modes. In this paper a wireless sensor network (WSN), comprised of purpose built resistive temperature sensor (RTS) nodes, is used in an MG distribution network to measure the line resistance, in order to identify illegitimate user(s) in real time. The changing resistance of a copper wire with the changing temperature in the sensing unit of an RTS node determines the real time resistance, which is then compared with the line resistance estimated through singular value decomposition at the sink node. In our experiments, it detects non-technical losses or theft in real time with 100% accuracy without knowing the exact topology of the distribution network.