Optimal Bidding Strategy Of Power Sales Company Under New Electricity Reform Environment

With the release of the new round of power reform "No.9 Document" in 2015,China's power system reform has entered a new stage.The document proposes to orderly open the power sales side of the power market,so that the price of electricity is formed by the market,and effectively play a decisive role in the allocation of resources in the market.The release of the electricity sales side means that power generation companies and social capital that do not meet the requirements can invest and set up sales companies to participate in competitive sales.Therefore,it is of great guiding significance to study how various new forms of electricity sales companies can better participate in the power market competition and formulate a more fair and efficient bidding strategy for the healthy operation of the electricity sales side and the electricity market.After the opening of the electricity sales market,there are many types of power sales companies.The research objects in this paper are divided into three categories,one is a single independent sales company,and the other is a combined power sales company that combines various power sales entities in the form of virtual power plants.Another type is a power sales company established by a traditional thermal power company.For an independent single power selling entity,the power purchase is first determined,and the wavelet load neural network model based on "meteorological similar day" is used to predict the user load demand.Secondly,in order to guide the sales company's quotation,a GA-BP neural network model based on fuzzy clustering is established to predict the short-term market clearing price.Finally,the Bayesian game model of the bidding company's bidding is constructed,and the optimal bidding strategy is obtained by obtaining the Bayesian Nash equilibrium solution.With renewable energy becoming the main direction of global energy development in the future,the emergence of virtual power plants provides an important way for distributed energy resources to participate in electricity market transactions.For this type of combined sales company,this paper applies multi-agent technology and Stackelberg dynamic game theory to design a two-layer nested dynamic game bidding model including virtual power plant and its internal distributed energy.Firstly,the internal market game bidding of virtual power plants with distributed energy is modeled.Secondly,the Bayesian game bidding model is established for multi-virtual power plants to participate in the external power market competition.Two layers of repeated dynamic games until the optimal equilibrium solution obtained.Although China has developed rapidly in the development and utilization of distributed energy in recent years,the problem of "abandoning wind and abandoning light"brought about at the same time has become increasingly serious.Affected by the renewable energy quota system and the green card trading policy,it is particularly important for a power selling company established by a traditional thermal power producer to formulate a bidding strategy so that it meets both policy requirements and maximization of benefits.Therefore,this paper studies the optimal quotation strategy of power generation companies established by power producers in the context of new energy policies,and establishes a two-layer optimization model to solve the problem.The first layer accounts for the green card transaction cost and establishes A profit maximization model of a power sales company based on competitors' quotations.The second layer constructs a minimum cost model for the transmission plug based on the node price method,and uses Monte Carlo simulation and genetic algorithm to solve the problem.