Day-ahead optimal scheduling of district-level integrated energy system considering data centre

摘要

With the rapid development of the information technology industry, data centres (DCs) spring up in recent years, deeply impacting the traditional energy system. This paper proposes a novel day-ahead optimal scheduling method for district-level integrated energy system (DIES) considering the integration of DCs. Firstly, all components in the DIES are modelled specifically, including energy station, network and DCs. Then, the optimal scheduling model is constructed aimed at minimizing the comprehensive cost. Constraints of energy purchase, energy station, energy subsystems and DCs are all contained, where different DCs can cooperate to deal with the computing workloads. The initial problem is finally simplified to a second-order cone programming (SOCP), which can be solved directly by Gurobi. Case studies are carried out on a test DIES which aggregates three DCs. Results show that DCs can optimize the power flow by transferring the computing workloads through the Internet, and reduce the total cost.