The transactive energy market (TEM) can dynamically regulate the distribution system's balance of supply and demand

which is a hot spot of current research. With the increasing development of data centers (DCs) with high load-shifting flexibility

it has become an important TEM market player. However

there are few studies on the optimal operation of DCs in TEMs. A bi-level optimization model of networked DCs in multiple TEMs is proposed to narrow the research gap. First

considering that the DCs with a considerable power consumption capacity can affect the market clearing price

the players in the TEM are divided into DC- and conventional-type prosumers

both of which are analytically modeled. Second

a TEM model including DCs is developed

and the alternating direction method of multipliers (ADMM) is employed to clear the market in a distributed way to avoid the privacy leakage of end users. Third

a bi-level optimal scheduling model of networked DCs in multiple TEMs is proposed to allocate the workloads among DCs optimally. The operating behaviors of TEMs where DCs are located are included. Fourth

differential evolution and metamodel are employed to solve the proposed bi-level optimization problem efficiently. Simulation results show that the proposed method can increase the benefit of networked DCs

improve the renewable energy utilization rate

and elevate the total social welfare.