In large-scale power grids, the remote active-active mode of the control center can comprehensively improve the disaster recovery performance of the system and realize the integrated operation and maintenance of active and standby. However, the off-site multi-center model will increase the level of energy consumption, which will bring challenges to the realization of the current carbon peak and neutrality targets of the power grid. Moreover, the current load distribution strategy between control centers does not take into account energy costs, data transmission costs and queuing delays. In this paper, the control center is considered to be directly powered by renewable energy sources, and a novel workload management framework is proposed to understand data transmission costs and queuing delays through intelligent scheduling decisions, taking a holistic approach to address cost minimization and renewable energy The problem of consumption is to achieve carbon peak and neutrality targets. First, the workload allocation problem is formulated as a non-cooperative game, and a Nash equilibrium-based intelligent game load allocation framework is designed to minimize operating costs. Second, the framework considers both energy and network cost minimization while meeting workload performance goals. Then, leverage detailed models to obtain a comprehensive set of characteristics that affect operating costs and workload performance, including throttling data center computing and cooling power, colocation performance disturbances, time-of-use tariffs, renewable energy, net metering, peak demand, throttling data Center queuing delays, and the costs involved in regulating the transfer of data between data centers. Finally, simulations show that the proposed method can reduce operating costs more effectively than existing methods, verifying its effectiveness.