Flexible Energy System Scheduling Based on the Coordination of Sodium-Ion Batteries and Data Centers in Extremely Cold Regions

The increasing penetration of renewable energy and rising computing demands intensify the need for flexible regulation resources. In extremely cold regions, low temperatures degrade energy storage and adjustable-load performance, challenging integrated energy systems (IESs). This study proposes a sodium-ion battery–data center coordinated scheduling framework tailored to seasonal cold-region operations. A unified model captures renewable generation uncertainty, temperature-dependent sodium-ion battery behavior, and computing load flexibility. A hierarchical mechanism is adopted, in which the upper layer employs an improved NSGA-III (I-NSGA-III) and the lower layer executes dynamic operations via fuzzy logic control. Case studies for winter and summer show operating cost reductions of 11.9% and 27.4%, renewable energy absorption rates of 90.87% and 86.55%, and loss-of-load rates of 1.28% and 0.53%, respectively. Compared with lithium-ion batteries, sodium-ion batteries further reduce the loss-of-load rate by 38.8%, improve absorption by 1.9%, and reduce operating costs by 10.7% under low temperatures. Unlike existing studies, this work integrates temperature-dependent battery behavior, constrained data center flexibility, and adaptive optimization–control coordination within a unified scheduling framework for cold regions. The proposed framework provides a practical engineering solution that improves economic performance, renewable energy utilization, and power supply reliability.