APPARATUS FOR CUSTOMER CLUSTERING USING AUTOMATIC METERING DATA AND METHOD THEREOF

摘要

PURPOSE: An apparatus for customer clustering using automatic metering data, and a method thereof are provided to minimize the region preponderance phenomenon by repetitively performing a process for dividing a load curve into a plurality of small clusters. CONSTITUTION: A load calculator(20) calculates daily average load curve from the average power usage of a specific time based on the metering data. A pre-processor(30) normalizes the daily average load curve. A clustering unit(50) repetitively performs small scale clustering about the cluster when the number of clusters about the normalized daily average load curve meets the final number of the cluster. The clustering part generates representative load pattern of the final cluster when the number of cluster meets the final target number.