首页|期刊导航|综合智慧能源|计及差异化需求响应的算电协同优化运行

计及差异化需求响应的算电协同优化运行OA

Optimal operation of computing-electricity coordination considering differentiated demand response

中文摘要英文摘要

针对高比例新能源接入下算电协同调节需求增强,新能源出力波动与数据中心算力负荷增长提高了系统柔性调节要求,而多主体差异化需求响应在滚动纠偏和结算校核方面考虑不足的问题,提出了一种计及差异化需求响应的算力-电力协同优化运行方法.将换电站和数据中心等不同的差异化负荷柔性能力划分为本地削负荷、时间平移和空间迁移3类需求响应服务,分别实现事件期负荷削减、可延期任务转移和跨区域负荷重分布,并构建市场化接口模型;建立节点响应价格模型,将电压风险、支路拥塞和峰荷裕度等因素纳入价格信号,实现需求响应出清结果的物理可行性校核;同时结合潮流与安全约束校验节点响应量,避免引发电压越限和支路重载.在此基础上,设计交替方向乘子法分布式出清方法,在保留库存、任务、服务等级协议、荷电状态和服务质量等本地约束的同时,通过公共变量交换实现多主体协调,结合实测响应、残差状态和恢复风险提出滚动价格纠偏机制.通过不同策略与中压和高中压配电网场景的算例结果表明,所提方法能够有效降低响应缺额,改善电压越限和网损水平,提高需求响应履约质量,增强动态执行过程中的安全性、经济性和结算稳定性,使差异化柔性资源转化为可采购、可交付、可结算和可校核的市场化响应服务,中压算例验证了不同需求响应服务组合及算法环节的机制有效性,高中压算例则检验了该方法在多电压等级、跨分区迁移条件下的工程适用性.为高比例新能源条件下算力负荷与配电网协同运行提供支撑.

The increasing integration of high-proportion renewable energy has intensified the demand for the regulation of computing-electricity coordination.Meanwhile,the fluctuations in renewable energy output and the growth of data center computing loads have imposed higher requirements for system flexibility,while existing differentiated demand response approaches among multiple stakeholders lack adequate consideration of rolling correction and settlement verification.To address these issues,an optimal operation method for computing-electricity coordination considering differentiated demand response was proposed.The flexible capabilities of differentiated loads,such as battery swapping stations and data centers,were classified into three types of demand response services:local load curtailment,temporal shifting,and spatial migration.These services enabled load reduction during events,deferrable task transfer,and cross-regional load redistribution,respectively.Additionally,a market-oriented interface model was constructed.A nodal response price model was established,with factors such as voltage risk,branch congestion,and peak-load margin incorporated into price signals to verify the physical feasibility of demand response clearing results.Nodal response quantities were also verified using power-flow and security constraints to prevent voltage limit violations and branch overloading.On this basis,a distributed clearing method based on the alternating direction method of multipliers was designed.While preserving local constraints on inventory,tasks,service-level agreements,state of charge,and quality of service,the method achieved coordination among multiple stakeholders through the exchange of common variables.A rolling price correction mechanism was proposed based on measured responses,residual states,and recovery risks.Case study results under different strategies and in both medium-voltage and medium-to-high-voltage distribution network scenarios showed that the proposed method effectively reduced response shortfalls,mitigated voltage limit violations,reduced network losses,improved demand response fulfillment quality,and enhanced safety,economic performance,and settlement stability during dynamic execution.The method transformed differentiated flexible resources into market-oriented response services that can be procured,delivered,settled,and verified.The medium-voltage case validated the effectiveness of different combinations of demand response services and algorithmic stages,whereas the medium-to-high-voltage case examined the engineering applicability of the method under multiple voltage levels and cross-zone migration conditions.The findings provide support for the coordinated operation of computing loads and distribution networks under high-proportion renewable energy conditions.

张冬冬;刘一宁;陈拓宇;武晓栋;江美慧;朱虹谕;张杰

内蒙古工业大学 新能源学院,内蒙古 鄂尔多斯 017010||内蒙古工业大学 风能太阳能利用技术教育部重点实验室,呼和浩特 010051||内蒙古工业大学 内蒙古自治区新能源与储能技术重点实验室,呼和浩特 010051||内蒙古工业大学 大规模储能技术教育部工程研究中心,呼和浩特 010051内蒙古工业大学 新能源学院,内蒙古 鄂尔多斯 017010湖南大学 电气与信息工程学院,长沙 410082内蒙古工业大学 新能源学院,内蒙古 鄂尔多斯 017010||内蒙古工业大学 风能太阳能利用技术教育部重点实验室,呼和浩特 010051内蒙古工业大学 新能源学院,内蒙古 鄂尔多斯 017010||内蒙古工业大学 风能太阳能利用技术教育部重点实验室,呼和浩特 010051内蒙古工业大学 新能源学院,内蒙古 鄂尔多斯 017010||内蒙古工业大学 风能太阳能利用技术教育部重点实验室,呼和浩特 010051||内蒙古工业大学 大规模储能技术教育部工程研究中心,呼和浩特 010051内蒙古工业大学 风能太阳能利用技术教育部重点实验室,呼和浩特 010051||内蒙古工业大学 内蒙古自治区新能源与储能技术重点实验室,呼和浩特 010051||内蒙古工业大学 大规模储能技术教育部工程研究中心,呼和浩特 010051||北京邮电大学 集成电路学院,北京 100876

能源科技

数据中心算电协同需求响应差异化建模交替方向乘子法

data centercomputing-electricity coordinationdemand responsedifferentiated modelingalternating direction method of multipliers

《综合智慧能源》 2026 (8)

57-68,12

国家自然科学基金项目(52677064)国家重点研发计划项目(2026YFE0211200)内蒙古自治区自然科学基金项目(2026QC0364,2026QC0321)National Natural Science Foundation of China(52677064)National Key R&D Program Project(2026YFE0211200)Inner Mongolia Natural Science Foundation(2026QC0364,2026QC0321)

10.3969/j.issn.2097-0706.2026.08.005

评论