AMI Based Load Estimation for Distribution System State Estimation

For accurate real-time monitoring of a distribution system, power flow or state estimation (DSSE) methods are used. Unfortunately, the accuracy of these tools can be compromised due to inaccurate load estimation. This paper addresses this issue by proposing a comprehensive method to improve load estimation for DSSE. The proposed method uses the data from the Advanced Metering Infrastructure (AMI) as it provides the most up-to-date and comprehensive load data. As the first step, a new clustering method is proposed to identify the loads that have similar profiles. The method is specially tailored for this application – a large set of time-series load profile data. In the second step, a guideline for the placement of bellwether meters using AMI is proposed to get more frequent load data update from these meters. In the third step, a regression method is proposed first for near-real-time load estimation. Then, this estimation tool is used together with the up-to-date load data from the bellwether meters to get accurate estimate of the loads in near-real-time. Test results show the effectiveness of the proposed approach.