摘要
针对燃煤火力发电厂锅炉燃烧系统复杂,控制难度大等问题,通过数据在线治理、深度强化学习,建立基于大数据实时分析驱动的人工智能算法模型,开发了一套智能燃烧优化指导系统,优化指导锅炉燃烧相关系统运行,以提高锅炉燃烧全过程的安全性和经济性。系统基于价值网络、约束网络和策略网络,利用深度强化学习技术,构建火电燃烧模拟器,通过大数据的样本积累,对相同负荷下锅炉效率进行优化排序推荐,从而在实际运行时,推荐最优燃烧工况的参数调整方案,锅炉效率逐步提升,并趋于平稳。
To address the complexities and difficulties in controlling the boiler combustion system in coal-fired power plants,a smart combustion optimization guidance system was developed by using online data governance,deep reinforcement learning,and big data real-time analysis to establish an artificial intelligence algorithm model based on the value network,constraint network,and policy network.The system optimizes and guides the operation of boiler combustion-related systems to improve the safety and economic efficiency of the entire boiler combustion process.The system uses deep reinforcement learning technology to construct a coal-fired combustion simulator based on the value network,constraint network,and policy network.It uses big data sample accumulation to optimize and rank the boiler efficiency under the same load,and then recommends the optimal parameter adjustment scheme for the actual operation,gradually improving the boiler efficiency and stabilizing it.
作者
刘成成
马艳明
陈一平
莫红军
LIU Chengcheng;MA Yanming;CHEN Yiping;MO Hongjun(China Energy Nanning Power Generation Co.,Ltd.,Nanning 530317,Guangxi,China;Electric Power Research Institute,National Grid Hunan Electric Power Co.,Ltd.,Changsha 410007,Hunan,China)
出处
《工业锅炉》
2025年第1期6-9,共4页
Industrial Boilers
关键词
燃煤火力发电
大数据实时分析
智能燃烧优化指导系统
深度强化学习
锅炉效率
coal-fired thermal power generation
big data real-time analysis
intelligent combustion optimization guid-ance system
deep reinforcement learning
boiler efficiency
