Attack Detection and Measurement Correction in Distribution Systems with Non-Synchronized Measurements

Synchronized measurements are required in many algorithms for real-time detection of attacks in the distribution system. However, accurately synchronized metering units, such as phasor measurement units (PMUs), are currently expensive to be massively deployed for billing purposes in distribution systems. Instead, smart meters are used, which report metering data with a timestamp that is usually skewed from the reference time due to the natural drift of the crystal oscillator used in the meter's real-time clock (RTC). Hence, unsynchronized smart meter measurements limit the practical implementation of many attack detection algorithms that assume synchronized measurements. This paper addresses this issue and proposes a new method that can be utilized by a fraud detection algorithm presented in the literature. Hence, the algorithm presented in the literature is modified to work properly in practical scenarios, i.e., when there is a skew in the time stamp received from smart meters. The performance of the modified algorithm is compared to the original algorithm presented in the literature, and the results show the robustness and effectiveness of the proposed method.