Computation Of Bidding Strategies Nash Equilibrium And Allocation Of Security Cost In Power Market

Because of the nature of the electric power industry and its background, an oligopolistic competition market model can represent this restructured environment better. This situation makes it more fit to make use of game theory in analyzing the participants'bidding behavior in power market. Nash equilibrium is one of the most important concepts in game theory. It represents final result of competition in power market. So the algorithm for finding Nash equilibrium becomes an important analytical tool and a key point of the research works for power market. But it is difficult to take the transmission line constraints and demand side bidding into consideration. In this dissertation, an attempt is made to solve these problems. We develop effective calculating methods of Nash Equilibrium corresponding to three different bidding strategies. In the meanwhile, the research on allocation of security cost is also involved in this dissertation.First of all, we develop a new method to search for the Nash equilibrium (NE) of linear supply function (LSF) model with incorporating the demand side bidding and transmission constraints, and analyze the feasibility of demand side bidding. In this model, four bidding parameters of both the generators and the demand sides are considered. An iterative process is proposed to find NE based on the sensitivities of profits of participates to bidding parameters. Without the transmission constraint, the pure strategy NE of LSF can be attained. However the existence of the transmission constraints may result in the inexistence of pure strategy NE. For this case, this method can achieve the equilibrium which is least affected by the change of loads.Secondly, a novel method for calculating the pure strategy Cournot NE, with the transmission constraints taken into account, is suggested. This method computes NE in the transmission states space of a grid. Since the number of the transmission states caused by the competition of the generators is quite small, contrast to the past algorithms, this method possesses advantages of less computation burden and higher computation precision. This method can easily judge the value of the transmission line capacity limit, at which the pure strategy NE does not exist. If it exists, then the NE can be obtained by this method. So this method is more suitable for the analysis of larger power market.Subsequently, two properties about Cournot model with the transmission constraints are proved. The two properties show the characteristics of the profit function of a generator in a given transmission grid state and the characteristics of the expected profit function of a generator with the mixed strategies respectively. On the basis of the two properties, we present a new method, to compute the mixed strategy NE of Cournot model, in which an iterative algorithm is proposed to update the generators'output strategy vectors of generator outputs until the vectors are unchanged forever. Comparing with the past algorithms, the algorithm given in this dissertation is more suitable for the multiplayer game.Finally, the Aumann-Shapley value method is improved and applied to the security...