부산대학교 도서관

The demand response of data center is considered as an effective flexible method to absorb excessive renewable energy (RE) in the power system, especially for the data center with high-density hydrogen storages. Therefore, this article proposes a collaborative response framework considering regulations of data workload and hydrogen storage. It is expected to further improve the data center flexibility. However, this collaborative response is a high-order and non-convex optimization problem involving the multi-period and complex model. Considering the condition of uncertain RE supply, it is difficult for power optimization dispatch to realize the precise decision in this NP-hard problem. To address this issue, a global interval optimization (GIO) approach is developed. It is to obtain efficient and robust decisions in the collaborative response. This approach deduces an interval optimization model to achieve the global optimality of convergent solution. Furthermore, this article introduces a probability correction into the GIO approach. It is used for improving the decision accuracy in the actual renewable power scenarios. A case study based on the IEEE 24-bus system verifies superiority of the GIO approach. It further validates effectiveness of the collaborative response framework for RE utilization.