综合智慧能源

• •    

基于历史数据驱动的分布式鲁棒IES低碳优化模型

王梓璇, 郝禹, 刘星辰, 孙韦男, 刘琳, 张逸   

  1. 吉林省气象信息网络中心, 130100
    国网吉林省电力有限公司长春市双阳区供电分公司, 吉林 130021 中国
  • 收稿日期:2025-05-06 修回日期:2025-07-04

Low carbon optimization model based on historical data-driven distributed robust IES

  1. , 130100,
    , 130021, China
  • Received:2025-05-06 Revised:2025-07-04

摘要: 为了有效提升经济效益减少碳排放,推进氢能的高效利用,降低可再生能源出力的不确定性,提出了一种基于历史数据驱动的分布式鲁棒综合能源系统(Integrated Energy System, IES)低碳优化模型。首先,采用数据驱动的不确定性量化方法,基于可再生能源出力的历史数据构建1-范数与∞-范数联合约束的概率分布模糊集,通过搜索最恶劣场景概率分布,构建min-max-min三层分布式鲁棒优化框架,在保障鲁棒性的同时降低调度保守性;其次,集成阶梯式碳交易机制、电制氢(Power-to-Gas, P2G)技术通过分阶段碳价激励减排,实现系统内碳元素的闭环利用,兼顾环境效益与经济成本;最后,采用列与约束生成(C&CG)算法求解。结果表明,基于混合范数的模糊集相比于单一范数的模糊集成本平均降低1.6%,进一步降低了可再生能源出力保守性;同时,在合理划分阶梯区间的情况下,碳排放量与碳交易成本分别下降16.35%和22.35%,达到了碳排放和碳交易成本的相对平衡,验证了其在降低碳交易成本方面更具优势。所提出的模型为综合能源系统的规划与运行提供了理论支撑与实践参考。

关键词: 综合能源系统, 分布式鲁棒, 阶梯式碳交易, 可再生能源, 碳排放

Abstract: In order to effectively improve the economic efficiency to reduce carbon emissions, promote the efficient use of hydrogen energy, and reduce the uncertainty of renewable energy output, a low carbon optimization model based on historical data-driven distributed robust Integrated Energy System (IES) is proposed. Firstly, a data-driven uncertainty quantification method is used to construct a fuzzy set of probability distributions with joint constraints of 1-parameter and ∞-parameter based on the historical data of renewable energy output, and a min-max-min three-layer distributed robust optimization framework is constructed by searching for the worst-case scenario probability distributions, so as to reduce scheduling conservatism while guaranteeing the robustness of the system; Secondly, the integration of stepped carbon trading mechanism, Power-to-Gas(P2G) technology through the phased carbon price incentives to reduce emissions, to achieve the closed-loop utilization of carbon elements in the system, taking into account the environmental benefits and economic costs; Finally, the column and constraint generation (C&CG) algorithm is used to solve the problem. The results show that the fuzzy set based on the integrated paradigm reduces the daily dispatch cost by 1.6% on average compared with the fuzzy set with a single paradigm, which further reduces the conservatism of the renewable energy output; meanwhile, in the case of rationally dividing the stepped interval, the carbon emission and the carbon transaction cost are reduced by 16.35% and 22.35%, which achieves the relative balance of carbon emission and carbon transaction cost, and verifies the relative advantage of it in reducing the carbon transaction cost is more advantageous. The proposed model provides a theoretical support and practical reference for the planning and operation of integrated energy systems.

Key words: integrated energy systems, distributed robustness, cascade carbon trading, renewable energy, carbon emission