Abstract
This paper propose a Nash Stackelberg game based trading decision model of joint power market contain frequency/regulation/reserve for day ahead transaction to deal with the challenges brought by the insufficient peak shaving and frequency regulation capacity of a high proportion of renewable energy. This model utilizes Copula-CVaR to quantify the risk of revenue loss caused by the uncertainty of power generation and consumption. The model based double layer game of Nash Stackelberg and considering the total cost of regulation and the profits of multiple types of independent operating entities. So the proposed model is complex with the traditional model because it is not only requires the balance between the upper and lower level entities, but also requires the balance between multiple types of power supplies in the lower level. The rationality and effectiveness of the trading decision model is verified by the measured data of the renewable energy gathering area in northwest China. The calculation results indicate that the trading strategy not only breaks through the limitations of poor flexibility in the power market caused by insufficient grid synchronization machines, but also solves the development bottleneck of long investment payback period and low utilization rate of energy storage stations.