Knowledge Extraction and Data Mining for the Competitive Electricity Auction Market

The transition from a vertically integrated industry to a horizontally integrated open market system changes the operational planning activities of generation companies (GENCOs). This transition, along with the strategic bidding decision process that must be employed by a GENCO, changes the objective from cost minimization to profit maximization. This change requires considering not only the technical aspects of unit operation, such as capacity limits, but also information about other market participants and the volatility of market prices. These additional factors are significant, especially in an oligopolistic market, because they influence the amount of electricity bought and sold, thus affecting net profit. This paper proposes an approach that data mines historical and current market data. The context is a deterministic four-market-participant environment. This model uses an auction simulator for 120 time periods. Results suggest that the data mining approach be extended to the reduction of epistemic uncertainty in VaR/PaR inferences using information gap theory