现代预测和优化算法在脱硫系统运行中的应用  

Application of modern prediction and optimization algorithms in FGD systems

作　　者：曹建宗 林宸雨 陈文通 刘琦[2] 周权[2] 要亚坤 樊帅军 马彩妮 马双忱 CAO Jianzong;LIN Chenyu;CHEN Wentong;LIU Qi;ZHOU Quan;YAO Yakun;FAN Shuaijun;MA Caini;MA Shuangchen(Shenzhen Energy Baoding Power Company Limited,Baoding 072150,China;Department of Environmental Science and Engineering,North China Electric Power University,Baoding 071003,China)

机构地区：[1]深能保定发电有限公司,河北保定072150 [2]华北电力大学环境科学与工程系,河北保定071003

出　　处：《华电技术》2020年第3期59-66,共8页HUADIAN TECHNOLOGY

摘　　要：人工智能技术的不断发展,为定量分析脱硫系统各因素变化带来的影响,从而预测系统脱硫效率和经济成本创造了条件。分析了近年来利用神经网络算法预测脱硫系统SO2排放量的算法网络结构、输入参数选择等算法细节,并讨论了神经网络算法应用于脱硫系统的特点。此外,针对基于脱硫系统设计的预测+优化算法的设计细节和优化目标,对增压风机、氧化风机和浆液循环泵的运行优化做了针对性讨论。The continuous development of artificial intelligence technology has impacted quantitative analysis flue gas desulfurization(FGD)system in various factors,and has created favorable condition for its desulfurization efficiency and economic cost.By analyzing the details of the network construction and parameter selection in FGD system SO2 emission prediction calculated by neural network algorithm,the application characteristics of the algorithm in FGD system is discussed.In addition,targeted discuss is made on the optimization of booster fan,oxidation fan and syrup circulation pump,which aims at refining the detailed design and optimization targets of the prediction+optimization algorithm in FGD systems.

关 键 词：脱硫系统 SO2排放量 预测 神经网络 优化 遗传算法 粒子群优化算法 智慧环保 人工智能 

分 类 号：X701.3[环境科学与工程—环境工程]