Research on Demand Response Optimization Strategies for Data Center Clusters from the Perspective of Multi-Spatio-Temporal Collaboration

To address the problems of high energy consumption in data center clusters and the neglect of the spatiotemporal synergy potential of load by traditional scheduling strategies, a multi-spatio-temporal synergy optimization strategy taking into account the load response characteristics of batch processing is proposed. By constructing a business flow-energy flow coupling model, server resources are dynamically configured, and multiple time Windows are divided based on the timeliness of the load, and spatio-temporal migration rules are established. With the goal of minimizing total operating costs and the constraint of user satisfaction, a mixed integer programming model is constructed. The simulation results show that this strategy can reduce operating costs, increase Central Processing Unit (CPU) utilization and user satisfaction, verifying the effectiveness of spatio-temporal co-scheduling.