Large Users’ Power Purchase Trading Strategy Method Based On Game Theory

In recent years, with the power market steady reform and deep perfecting, demand side is in constant open. In order to avoid the monopoly of the traditional power grid, the market gradually allows large users to choose the proper power suppliers freely. The moment directly purchasing power with a self-supply power plant, large users can select more objects, and both sides can also gain their expected optimal profit through the mode of centralized bidding. Therefore, it’ll be important to study the direct power purchase trading strategies joined by big users, and the strategies have a vital significance in perfecting competition system of electric markets and optimizing the ultimate profit of bilateral trade issues.A new strategy about the optimal priority of the bilateral trade is proposed by this thesis, and based on the existing talk on large user electricity purchasing, we solve the market price and quantity selection problems between the large users and power suppliers in the process of trading, which takes the probability distribution of game model into consideration This strategy transfers the complex transaction mode into a competitive one in which large users firstly determine the electricity price and quantity priority. Using the first sealed-hid bidding model, we can analyze how they maximize the expected purchase price to obtain the optimal choice rights for the self-supply plant and ensure the profit optimization at the same time. An untruth probability distribution issue is put forward to forecast the electricity price competition system more practically and accurately, when the large users purchase the power.In addition, as for the prediction issue for power, the grey forecasting model is indicated to predict the load faced by large users, and also comes up with a new power predication model between large users. This mode brings in a virtual opponent, which transfers multi-game into a double game, and solve the forecast issues by using the theory of repeated game algorithms. The thesis innovatively points out the users’ power schedule strategy, and influences the electricity pricing during the peak period, so as to get the purpose of reducing peak and filling valley. Meanwhile the users can obtain the optimum power purchase expenses. Rubinstein model is introduced in the thesis to design the bilateral trading strategy on the basis of above discussion, and Bayesian-mixed strategy is also designed. The strategy for the power market bilateral exchange is chosen through the data simulation. In the transaction process among more large users and more power generating company, the large user can better the profit by considering the strategy of opponents’ probability distribution, update strategy selection and achieve the transaction price more quickly.Finally, CVaR is used to study the risk value in the process of trade in the thesis. It provides large users with the power purchase transactions against risks, for a better strategy.