基站
需求响应
热的
基础(拓扑)
环境科学
计算机科学
动态需求
负荷管理
工程类
控制理论(社会学)
火力发电站
频率响应
基本负荷发电厂
温度测量
功率(物理)
汽车工程
响应时间
作者
Yu Guo,Tianguang Lü,Yansong Mei,Haokun Liu
标识
DOI:10.1109/tsg.2026.3751494
摘要
The increasing uncertainty of renewable generation and loads highlights the need for distributed flexibility, and large-scale 5G base station clusters (BSCs) are promising demand-response (DR) resources. However, existing studies mainly focus on cost-oriented scheduling and lack network-feasible quantification of bidirectional DR capacity under coupled communication and energy constraints. This paper proposes a multidimensional, network-feasible DR quantification framework that jointly integrates the flexibility of communication load (CL), air-conditioning (AC), and backup energy storage (BES) across the communication and distribution networks. Dynamic user–BS reassociation and CL migration are characterized by a time-varying bipartite graph, while thermal and reliability-aware models quantify AC and BES flexibility; the resulting spatial BSC power variations are further embedded into a radial distribution network through LinDistFlow-based voltage and branch constraints, thereby directly coupling communication-side flexibility reconfiguration with grid-side operating feasibility. A GNN-guided stage-wise algorithm is further developed to enable the solution of the resulting large-scale optimization problem. Simulation results show that, while satisfying the communication and distribution-network operating constraints, the proposed framework increases the average cluster power by approximately 110.9% under DR-Up and reduces it by approximately 98.8% under DR-Down relative to the baseline. The proposed GNN-guided stage-wise method also reduces the computation time by up to 66.4% compared with the benchmark methods.
科研通智能强力驱动
Strongly Powered by AbleSci AI