基于联邦ADMM数据隐私保护的电-氢耦合主动配电网分布式协同优化运行OA
Distributed collaborative optimal operation for electricity-hydrogen coupled active distribution networks based on federated ADMM for data privacy protection
针对高比例分布式电源接入下主动配电网新能源消纳不足、电压频繁越限、数据隐私风险高及计算效率低等问题,提出一种基于联邦交替方向乘子法(federated alternating direction method of multipliers,FedADMM)的隐私保护分布式优化算法.首先,建立FedADMM算法数学模型,通过避免子问题闭式求解提高计算效率.其次,构建含电氢耦合系统的配电网运行优化模型,设计基于 FedADMM 的三层分布式优化求解框架.最后,在 IEEE33和 IEEE123 节点系统上进行仿真.结果表明,所提方法使系统新能源消纳水平提升 7.82%,较 ADMM 计算效率提升80.36%,较基于同态加密的ADMM具有更高计算精度.
To address the insufficient renewable energy accommodation,frequent voltage violations,high data privacy risks,and low computational efficiency in active distribution networks with high penetration of distributed generation,a privacy-protection distributed optimization algorithm based on the federated alternating direction method of multipliers(FedADMM)is proposed.First,a mathematical model of the FedADMM algorithm is established,in which computational efficiency is improved by avoiding the closed-form solutions of subproblems.Second,an optimization model for the electricity-hydrogen coupled distribution network is developed,and a FedADMM-based three-layer distributed framework is designed.Finally,simulations on the IEEE33-bus and IEEE123-bus systems show that the proposed method increases renewable energy accommodation by 7.82%,improves computational efficiency by 80.36%over ADMM,and achieves higher accuracy than homomorphic encryption-based ADMM.
顾梨婷;崔明建;田云鹏;邓芳明;康兵;卢明;黄小庆;陈轲娜
天津大学电气自动化与信息工程学院,天津 300072天津大学电气自动化与信息工程学院,天津 300072辽宁工程技术大学电气与控制工程学院,辽宁 葫芦岛 125105华东交通大学电气与自动化工程学院,江西 南昌 330013江西水利电力大学电气工程学院,江西 南昌 330099湖南科技大学信息与电气工程学院,湖南 湘潭 411201湖南大学电气与信息工程学院,湖南 长沙 410082国网四川省电力公司电力科学研究院,四川 成都 610041
FedADMM隐私保护电氢耦合系统分布式优化
FedADMMprivacy protectionelectricity-hydrogen coupled systemdistributed optimization
《电力系统保护与控制》 2026 (18)
13-24,12
This work is supported by the National Natural Science Foundation of China(No.52207130). 国家自然科学基金项目资助(52207130)江西省自然科学基金项目资助(20253BAC260014)
评论