基于核能供热技术的热网调控与智慧化管理
王锋¹，孙英策¹，刘慧文²，仇天星¹
（1. 国家电投集团东北电力有限公司大连大发能源分公司，辽宁 大连 116000；
2. 内蒙古工业大学电力学院，内蒙古 呼和浩特 010051）
摘要：核电供热技术凭借其显著的低碳和高效特性，已成为实现“双碳”目标的重要手段之一。然而，当前核
能供热负荷预测不准确、调控手段单一，无法满足该技术发展需求。为此，提出了一种基于卷积循环注意力机制
模型（CGAM） 的精细化调控和热负荷预测方法用于核电供热系统。该方法首先采用多层次数据存储技术和数据
同步与优化策略，确保生产数据的实时性与准确性；其次引入深度网络对核电换热站短期热负荷进行精确预测并
实施策略调整，确保预测精度；最后依据预测结果设置多层次精准调控手段，完成供热系统的智能调节。将该方
法应用于大连红沿河核电供热项目中，实验结果显示，对实际工况的热负荷实现了准确预测，对供热系统起到了
精准调控的作用，证明了算法的有效性。
关键词：核电供热；精细化调控；热负荷预测；智慧化管理
中图分类号：TM623      文献标志码：A      文章编号：2095-0802-(2024)01-0318-07
Heating Network Regulation and Intelligent Management Based on Nuclear
Heating Technology
WANG Feng1, SUN Yingce1, LIU Huiwen2, QIU Tianxing1
(1. Dalian Dafa Energy Branch Company, Northeast Electric Power Co., Ltd., State Power Investment Group, Dalian
116000, Liaoning, China; 2. School of Electric Power, Inner Mongolia University of Technology, Hohhot 010051, Inner
Mongolia, China)
Abstract: With its remarkable low-carbon and high-efficiency characteristics, nuclear power heating technology has
become one of the important means to achieve the goal of "dual carbon". However, the current nuclear heating load
prediction is inaccurate and the control means are single, which cannot meet the needs of this technology development.
Therefore, a refined regulation and heat load prediction method based on convolutional cycle attention mechanism model
(CGAM) was proposed for nuclear power heating system. Firstly, multi -level data storage technology and data
synchronization and optimization strategy were adopted to ensure the real-time and accuracy of production data. Secondly,
a deep network was introduced to accurately predict the short-term heat load of the nuclear power heat exchange station
and implement strategy adjustment to ensure the prediction accuracy. Finally, multi-level precise control means were set
according to the prediction results to complete the intelligent adjustment of the heating system. The method was applied to
Dalian Hongyanhe Nuclear Power Heating Project. The experimental results show that the heat load of the actual working
conditions is accurately predicted, and the role of precise regulation of the heating system is achieved, which shows the
effectiveness of the algorithm. Key words: nuclear power heating; fine regulation; heat load forecasting; intelligent management