Self-organizing classification and identification of miscellaneous electric loads

Miscellaneous electric loads (MELs) represent a large portion of the electricity consumption. Economic and environmental impacts of energy consumption lead to needs and opportunities in energy management and saving. This paper proposes an intelligent classification and identification of MELs by extending the Self-Organizing Map (SOM) framework to a supervised manner. The SOM can classify a large amount of MELs data into several clusters by inherent similarities. The self-organizing identifier thus has the advantages of being accurate, robust, and applicable.