The Method Of Bidding In Power Market Based On Immune Genetic Algorithm

With the merging of the power market in China, power plants were faced with a serious challenge. The power market on generation-side of separating power plants from power networks and generation bidding will become the main operating model in our hometown. It is necessary to improve social benefits and reduce the purchasing price with the mature power technology. This thesis researches on bidding method of power market, according to the development direction of power market in our country.At first, this thesis analyzes the emergent problem needed to be solved in power market and gives some background about power market; secondly, generation cost curve has been analyzed in this thesis, which generation cost is primary part of electrical price; and then, the bidding models and principle of power market has been described detailed; at last, the thesis presents a bidding method based immune genetic algorithm.At present, there are so many types of generation bidding algorithms commonly used in deregulation environment, like the equal bidding price method, merit-order method, the dynamic programming method and so on. But they have big limitation, because they can be used to solve different bidding curves respectively and deal with various sets of constraints. Immune algorithm is a immune system simulating biological immune system, in which the antigen and antibodies are categorized as the objective function and the solution respectively. The antibody that combined perfected with the antigen is deemed the optimum solution to the problem. One immune genetic algorithm has been explored for bidding of power market in this thesis, which opens up one new kind of approach to solve the problem of bidding after the encoding style, immune operator, the method to get the immunity and repulsion, the choosing method of antibody and the renewing method of the antibody memory bank have been confirmed in this thesis.The typical computational example shows that presented algorithm is a new effective optimal method, as it can obtain better result and has faster convergence speed, and also it has approximately linear computational complexity.