Analysis Of Bidding Strategy For Generation Company And Compute CVaR Based On Copula Function

The Copula theory has advantages applied to the relevant analysis in financial market. In the electric power market's risk analysis, we may decompose the property risk into a single asset risk and portfolio risk. The single property's risk can be described by their boundary distribution , while the profolio produces the risk completely described by Copula function. Considering the principle of power market, we establishment risk model.Utilizing the Monte Carlo simulation method to copute the CVaR based on Copula Function. The primary contents are as follows:In the first section, we mainly introduce the risks management in electricity market ,In addition, topic background and our research works of this paper are also briefly introduced.In the second section,we introduce the Copula function and the CVaR definition as well as the correlation theorem.In the third section, Through the historical price data's processing, this paper simulates the uncertainty factor in the day ahead market and the contract market to the lognormal distribution , then the revenue functions of the power generation companies in the day ahead market and the contract market are obtained respectively. Based on this establishment the bidding model considering contract. Due to the relevance of the two market revenue, Gumbel Copula function is chosen and it is applied to the risk measure of power market. Utilizing the Monte Carlo Simulation method to compute the CVaR.Meanwhile,this paper analyses the result and obtains the optimal bidding strategy of power generation companies:The power generation companies with better risk tendency will put very small part of electricity into the contract market.They prefer a high bidding strategy.While power generation companies avoiding risk,will increase the quantities of the electricity in the contract market,and they prefer a lower bidding strategy.In the fourth section, we introduce three types of portfolio optimization models, Meanwhile,we set general model of the maximal profit with CVaR. Through the historical price data's processing, utilizing the Monte Carlo Simulation method to compute the CVaR and obtain effective front of this model, the conclusion also comply with the market behavior.