Data mining techniques for electricity customer characterization

The liberalization of electricity markets has been resulted in the emergence new players, increasing competitiveness markets, standing those can provide better services for prices. knowledge energy consumers’ profile an important tool to help players make decisions electrical sectors. In this paper, a characterization model typical load curves Low Voltage (LV) customers is proposed and evaluated. identification consumption patterns based on clustering analysis. methodology seven algorithms, partitional hierarchical. Also, five validity indices are used identify best data partition. With obtained analysis, classification classify according their data. select correct class each customer. To simple, curve represented by three which represent shape. work demonstrates be effective most diverse sectors, highlighting use optimization contracting low voltage customers. constantly updated improve precision, finding estimates that consumers habits.