Data mining for prosumers aggregation considering the self-generation

Abstract

Several challenges arrive with electrical power restructuring, liberalized electricity markets emerge, aiming to improve the system’s efficiency while offering new economic solutions. Privatization and liberalization of previously nationally owned systems are examples of the transformations that have been applied. Microgrids and smart grids emerge and new business models able to cope with new opportunities start being developed. New types of players appear, allowing aggregating a diversity of entities, e.g. generation, storage, electric vehicles, and consumers, Virtual Power Players (VPPs) are a new type of player that allows aggregating a diversity of players to facilitate their participation in the electricity markets. A major task of VPPs is the remuneration of generation and services (maintenance, market operation costs and energy reserves), as well as charging energy consumption. The paper proposes a normalization method that supports a clustering methodology for the remuneration and tariffs definition. This model uses a clustering algorithm, applied on normalized load values, the value of the micro production, generated in the bus associated to the same load, was subtracted from the value of the consumption of that load. This calculation is performed in a real smart grid on buses with associated micro production. This allows the creation of sub-groups of data according to their correlations. The clustering process is evaluated so that the number of data sub-groups that brings the most added value for the decision making process is found, according to players characteristics.