Distributed privacy-preserving optimization for multi-energy system considering carbon emission

Multi-energy systems (MESs) present great potential for reducing operational costs and carbon emissions. Nodal carbon intensity (NCI) can effectively quantify the carbon emissions at the nodal level. The NCI's potential for carbon emission reduction is yet to be unlocked on the demand side. However, the nonconvexity of the NCI function poses a serious computational challenge, and the corresponding large-size problem is unsolvable even for state-of-the-art solvers. To bridge these gaps, this work introduces an optimization model to reduce carbon emissions via demand management and proposes a novel successive approximation algorithm to address its notorious computational challenge. On the other hand, stakeholders in MESs often have different interests. There is thus a strong demand to manage energy in a distributed fashion and to preserve privacy in MES. A novel information-exchange mechanism and acceleration technique for distributed optimization are proposed based on homomorphic encryption (HE), achieving privacy protection and fast computational performance. Case studies validate the effectiveness and performance of the proposed approaches.