双碳目标下电网异址双活调控中心负载均衡控制方法

在大型电网中，调控中心异址双活模式可全面提升系统容灾性能，实现主备一体化运维。然而异址多中心模式会增加能源消耗水平，给当前电网双碳目标的实现带来挑战。目前调控中心间负载分配策略并没有综合考虑能源成本、数据传输成本和排队延迟。调控中心考虑由可再生能源直接供电，并提出了一种新颖的工作负载管理框架，通过智能调度决策了解数据传输成本和排队延迟，采用整体方法来解决成本最小化和可再生能源消纳问题，即实现双碳目标。将工作负载分配问题表述为非合作博弈，并设计了一种基于纳什均衡的智能博弈负载分配框架，以最小化运营成本。该框架同时考虑能源和网络成本最小化，且同时满足工作负载性能目标。利用详细的模型来获得影响运营成本和工作负载性能的一组综合特征，包括调控数据中心计算和冷却功率、托管性能干扰、分时电价、可再生能源、净计量、高峰需求、调控数据中心排队延迟，以及调控数据中心间数据传输所涉及的成本。通过仿真表明，所提出的方法可以比现有方法更有效地降低运营成本，验证了其有效性。

Load balancing control method for remote active-active control centers of power grid with carbon peak and neutrality targets

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.