Multi-level games optimal scheduling strategy of multiple virtual power plants considering carbon emission flow and carbon trade

Under the goal of "carbon emissions peak, carbon neutrality", virtual power plant (VPP) is of great significance in improving grid safety level and promoting the clean and low carbon energy transition. However, there is a significant contradiction between the weather-dependent and intermittent output of renewable energy and the combined heat and power (CHP) unit working in the way of "with heat to determine electricity" in winter heating areas. It will seriously affect the peak-load regulating capacity of VPP, leading to high carbon emissions. With the gradual deepening of low-carbon energy transition and the continuous improvement of the carbon market, it provides a possible approach for solving the above problem. Therefore, this paper proposes a multi-level games optimal scheduling strategy of multiple virtual power plants considering carbon emission flow and carbon trade. The structure of VPP built in this paper adds carbon capture and storage (CCS), electrical energy storage device and electric boiler on the basis of CHP and renewable energy power generation, and considers the carbon-oriented demand response mechanism. The multiple VPPs architecture is established that can meet the demands of clean heating and energy supply. Then, the model of carbon emission flow (CEF) suitable for VPP structure in this paper is established. On the basis, a multi-level games optimal scheduling model is established. The Nash-bargaining model is used between multiple VPPs to simulate the gaming behavior of each VPP in the carbon trading market. Master-slave game is used within the VPP to guide the low carbon transformation in demand-side through the carbon-oriented price mechanism. Finally, adaptive alternating direction multiplier method (ADMM) combined with data-driven two-stage robust optimization is used to solve the model, in order to obtain the optimal trading volume of carbon quota and trading price. The parallel column and constraint generation (CCG) algorithm is used to increase the efficiency of model solution. The simulation results show that the algorithm of parallel CCG and adaptive ADMM has higher accuracy and shorter computing time. The proposed scheduling strategy can achieve flexible and low-carbon operation of VPP. Through coordination control between the source side and the demand-side, scheduling strategy can effectively help VPP improve the capacity of renewable energy utilization, reduce carbon emissions, the costs of VPP source side and demand-side.