| Resumo: | The relationship between generation and demand can be accomplished by centralized forward markets, as the day ahead pool market, or by bilateral contracts. In their simplest version, day ahead pool markets are bid based uniform price mechanisms that receive buying and selling bids from the market agents. These bids usually correspond to pairs of quantity / price, that is, available quantity of power, minimum price to receive for selling bids and quantity to be supplied, maximum price to pay for buying bids. If no more information is provided, the day-ahead market corresponds in fact to 24 hourly uniform auctions in the sense that each one is independent from the results in previous ones and does not determine or have any impact in subsequent ones. In fact, this simple model has to be enhanced by considering multi block generator bids and by admitting several types of constraints that couple the referred hourly dispatches. In this case, we have a set of complexity constraints that may turn unfeasible the set of 24 dispatches from the independent auctions. Together with possible violations of network constraints, this is why that initial schedule may have to be changed. In this paper, we formulate the multi-period auction problem and we detail the solution algorithm based on Simulated Annealing in order to overcome unfeasibilities related both with complexity and network constraints. The developed approach is tested using a Case Study based on the IEEE 24 bus / 38 branch test system.
|
|---|