Auction-Based Dispatch Algorithms in Deregulated Power Systems

In deregulated power systems, the classic economic dispatch has been replaced with the auction-based dispatch in which both the generating units and loads participate. Instead of constant load demand in the classic economic dispatch, the loads are variables in the new dispatch. As a result, the conventional economic dispatch algorithms no longer can be applied. In this paper we reformulate the auction-based dispatch into a general minimization problem so that the loads can be treated as variables similar to the generations. Based on this new reformulation, efficient algorithms to solve the auction-based dispatch problem are proposed. The first algorithm solves the dispatch problem in which the objective function contains only quadratic bidding functions. Since some loads might submit linear incremental bidding functions, a second algorithm is developed to handle the situation where the objective function contains both quadratic and linear incremental bidding functions. At the end, we demonstrate the efficiency of our algorithms through examples.