氢能数据中心的柔性负荷调度：一种效率优先的非凸不确定优化方法

The hydrogen energy data center (HEDC) represents an emerging clean energy infrastructure. Due to its potential load flexibility, HEDC is regarded as a valuable object for power demand response (DR) programs. This paper designs a DR-based scheduling strategy to optimize HEDC operations, achieving simultaneous improvement in electricity cost reduction and renewable energy (RE) utilization. This optimization strategy needs to address three key non-convex challenges: complex load characteristics of HEDCs, dynamic electricity pricing, and RE uncertainty. Focusing on these challenges, a non-convex set interval optimization (NSIO) approach is proposed to obtain high-performance and precise solutions under uncertain conditions. The advantages of the proposed NSIO are that: (i) modeling properties of efficiency preference; (ii) exact convex relaxation techniques for reducing computational complexity; (iii) fully explicit mathematical modeling for achieving global optimality. Comprehensive case studies are conducted on the modified IEEE 24-bus power system. The results validate that the above response strategy effectively improves operational costs and RE utilization. Through systematic evaluation of five state-of-the-art uncertain optimization approaches, the proposed NSIO demonstrates 2.3 % 20.0% superior performance compared to classical approaches.