Closed-loop prediction and optimization of multi-energy flow coupling in virtual power plants

In order to cope with global warming and control carbon emissions, virtual power plants (VPPs) have become a key means to improve the capacity of renewable energy consumption and reduce the ecological impact of existing energy infrastructure. This paper proposes a multi-energy flow coupled VPP scheduling optimization model based on electrical synergy and carbon capture technology combined with the organic Rankine cycle. In the model, waste heat boilers and heat pumps are used to improve the energy conversion efficiency and flexibility of the combined heat and power system and to achieve the coordinated scheduling and complementarity of electricity-, heat-, and gas-based multi-energy flows. In view of the complexity of the VPP system and the need for the coordinated scheduling of multiple units, a closed-loop prediction and optimization (CPO) framework is introduced to reduce the system operating cost by coordinating renewable-energy forecasting with multi-energy scheduling decisions. The effectiveness of the proposed model is verified by examples. Experimental results show that compared with the open-loop prediction and optimization (OPO) framework, the CPO framework significantly improves the economy of the VPP system, reduces the actual cost of the system by 2.75%, and improves the daily economy by 13.41%. Although the cost-oriented prediction accuracy is lower than that of OPO in some time periods, it effectively reduces the additional cost caused by rescheduling (SUSD, i.e., start-stop cost, is reduced by 53%), takes into account both economy and prediction accuracy in multi-energy flow coupling scheduling, and improves the overall efficiency of the system.