| 55 | 0 | 62 |
| 下载次数 | 被引频次 | 阅读次数 |
【目的】随着中国能源结构复杂化，火电机组深调运行时面临稳燃困难、能耗升高与污染物调控失准等问题。为提升火电机组负荷波动所导致的入炉燃料量、发热量、烟气成分及主蒸汽流量等关键参数动态变化时的测量准确度，克服传统测量手段存在周期长、成本高或依赖经验值的局限，【方法】本文构建了基于机理-数据混合驱动的软测量技术体系：通过热力学平衡方程与燃烧反应机理建立辅助变量关联网络；采用异常值剔除、时序对齐及主成分分析提升数据质量；结合支持向量机、随机森林等算法的非线性映射能力，嵌入滤波等动态补偿机制，形成自适应变工况的预测模型。【结果】研究表明，软测量技术应用于火电机组可实现入炉煤发热量预测误差<0.5 MJ/kg、NOx浓度测量延迟≤30 s、主蒸汽流量精度>98.5%，支撑660 MW机组供电煤耗降低1.1 g/(kW·h)、脱硝系统氨逃逸率下降35%。【结论】提出多场景建模准则，针对小样本/线性问题优选支持向量机模型，对强时序特性参数采用长短期记忆网络，为工程人员提供跨工况的算法选型依据。通过软测量技术对核心参数的精准感知与动态优化，可以提升火电机组在宽负荷条件下的控制品质与运行能效。
Abstract：[Objective] With the increasing complexity of energy structures in China, coal-fired power units operating under variable conditions face safety and economic challenges such as combustion instability, rising energy consumption, and inaccurate pollutant regulation. Fluctuations in unit load cause dynamic variations in critical parameters, including fuel feed rate, calorific value, flue gas composition, and main steam flow rate. However, traditional measurement methods suffer from limitations such as long cycle times, high costs, or reliance on empirical values. [Methods] To address these issues, this study establishes a hybrid mechanism-data-driven soft sensor framework. A correlation network of auxiliary variables is constructed based on thermodynamic equilibrium equations and combustion reaction mechanisms. Data quality is enhanced through outlier removal, time-series alignment, and principal component analysis. By integrating the nonlinear mapping capabilities of algorithms such as support vector machines(SVM) and random forests, along with embedded dynamic compensation mechanisms like filtering, an adaptive prediction model for variable operating conditions is developed. [Results] Case studies demonstrate that the application of soft sensor technology in coal-fired power units achieves a coal calorific value prediction errors below 0.5 MJ/kg, NOx concentration measurement delays under 30 s, and main steam flow measurement accuracy exceeding 98.5%, supporting a 1.1 g/(kW·h) reduction in coal consumption for a 660 MW unit and a 35% decrease in SCR system ammonia slip. Multiscenario modeling criteria are proposed. SVM is prioritized for small-sample/linear problems, while long short-term memory(LSTM) networks are recommended for parameters with strong temporal characteristics, offering engineers cross-condition algorithm selection guidelines. [Conclusion] Propose multi-scenario modeling criteria, prefer the SVM model for small-sample/linear problems, adopt the LSTM network for parameters with strong time-series characteristics, and provide algorithm selection criteria for engineers across loading conditions. Through precise sensing and dynamic optimization of core parameters via soft sensor technology, this study aims to significantly enhance the control quality and operational efficiency of coal-fired power units under wide-load conditions.
[1]侯儒伟，马华.基于PSO-Elman的火电厂入炉煤热值预测模型研究[J].能源研究与利用，2024(2):10-14+56.HOU Ruwei, MA Hua. Research on prediction model of coal calorific value in thermal power plants based on PSO-Elman[J].Energy Research&Utilization, 2024(2):10-14+56.
[2]ZHANG Y, HUTCHINSON P, LIEVEN N A J, et al. Remaining useful life estimation using long short-term memory neural networks and deep fusion[J]. IEEE Access, 2020, 8:19033-19045.
[3]HUANG C Z, HE J X, ZHENG W, et al. A health monitoring method based on multivariate-time series adaptive gated recurrent unit transfer learning model for coal mill system[J]. Reliability Engineering and System Safety, 2025, 256:110767.
[4]李泽铭,田亮.基于卡尔曼滤波器的循环流化床机组燃料发热量软测量[J].华北电力大学学报(自然科学版), 2021, 48(2):89-95+113.LI Zeming, TIAN Liang. Soft measurement of fuel calorific value for CFB units based on Kalman filter[J]. Journal of North China Electric Power University(Natural Science Edition), 2021, 48(2):89-95+113.
[5]李祥宇，隋璘，马君霞，等.基于时序迁移与双流加权的ONLSTM软测量建模[J].化工学报，2023,74(11):4622-4633.LI Xiangyu, SUI Lin, MA Junxia, et al. ONLSTM soft sensor modeling based on temporal transfer and dual-stream weighting[J].CIESC Journal, 2023, 74(11):4622-4633.
[6]李健，尚星宇，王瑞.基于聚类算法的双进双出磨煤机入炉煤量软测量方法[J].热力发电，2020,49(11):114-119.LI Jian, SHANG Xingyu, WANG Rui. Soft measurement method for coal feed quantity of double-inlet and double-outlet coal mill based on clustering algorithm[J]. Thermal Power Generation, 2020,49(11):114-119.
[7]WANG Y N, CHEN X, ZHAO C H. A data-driven soft sensor model for coal-fired boiler SO2 concentration prediction with nonstationary characteristic[J]. Energy, 2024, 300:131522.
[8]ZHANG X W, HAN L, MA T, et al. Non-linear soft sensingmethod for temperature of coal spontaneous combustion[J]. Process Safety and Environmental Protection, 2023, 170:1023-1031.
[9]李京.燃煤机组脱硫吸收塔入口二氧化硫浓度软测量方法研究[D].北京:华北电力大学，2021.LI Jing. Research on soft measurement method of SO2concentration at desulfurization absorber inlet of coal-fired units[D]. Beijing:North China Electric Power University, 2021.
[10]杨丽，吴雨茜，王俊丽，等.循环神经网络研究综述[J].计算机应用，2018,38(S2):1-6+26.YANG Li, WU Yuqian, WANG Junli, et al. Review of recurrent neural networks[J]. Journal of Computer Applications, 2018, 38(S2):1-6+26.
[11]俞金寿.软测量技术及其应用[J].自动化仪表，2008(1):1-7.YU Jinshou. Soft sensor technology and its applications[J]. Process Automation Instrumentation, 2008(1):1-7.
[12]FARIDI K I, TSOTSAS E, HEINEKEN W, et al. Development of a neural network model predictive controller for the fluidized bed biomass gasification process[J]. Chemical Engineering Science,2024, 293:120000.
[13]姚顺春，刘泽明，卢志民，等.软测量技术赋能燃煤电厂碳排放计量的研究进展[J].洁净煤技术，2024,30(8):18-31.YAO Shunchun, LIU Zeming, LU Zhimin, et al. Research progress of soft sensor technology empowering carbon emission measurement in coal-fired power plants[J]. Clean Coal Technology, 2024, 30(8):18-31.
[14]ALHAJERI M S, ALNAJDI A, ABDULLAH F, et al. On generalization error of neural network models and its application to predictive control of nonlinear processes[J]. Chemical Engineering Research and Design, 2023, 189:664-679.
[15]李俨，张乘铭，刁云鹏，等.煤质低位发热量软测量技术原理及应用[J].吉林电力，2021,49(3):34-37.LI Yan, ZHANG Chengming, DIAO Yunpeng, et al. Principle and application of soft sensing technology for low calorific value of coal quality[J]. Jilin Electric Power, 2021, 49(3):34-37.
[16]韩忠旭，周传心，李丹，等.燃煤发热量软测量技术及其在超临界机组控制系统中的应用[J].中国电机工程学报，2008,28(35):90-95.HAN Zhongxu, ZHOU Chuanxin, LI Dan, et al. Soft sensing technology of coal calorific value and its application in supercritical unit control system[J]. Proceedings of the CSEE, 2008, 28(35):90-95.
[17]杨耀权，张新胜. LSSVM动态软测量模型在磨煤机一次风量预测方面的应用[J].动力工程学报，2016,36(3):207-212+217.YANG Yaoquan, ZHANG Xinsheng. Application of LSSVM dynamic soft sensor model in primary air flow prediction of coal mill[J]. Journal of Chinese Society of Power Engineering, 2016, 36(3):207-212+217.
[18]林伟俊，李敏，赵畅，等.基于延迟特性的NOx浓度软测量与喷氨优化[J].动力工程学报，2023,43(10):1352-1362.LIN Weijun, LI Min, ZHAO Chang, et al. Soft measurement of NOx concentration and ammonia injection optimization based on delay characteristics[J]. Journal of Chinese Society of Power Engineering, 2023, 43(10):1352-1362.
[19]赵征，曾德良，田亮，等.基于数据融合的氧量软测量研究[J].中国电机工程学报，2005(7):7-12.ZHAO Zheng, ZENG Deliang, TIAN Liang, et al. Research on oxygen soft measurement based on data fusion[J]. Proceedings of the CSEE, 2005(7):7-12.
[20]杨耀权，张新胜.基于自适应变异SAPSO-LSSVM的磨煤机一次风量预测[J].自动化仪表，2016,37(7):88-92.YANG Yaoquan, ZHANG Xinsheng. Prediction of primary air flow in coal mill based on adaptive mutation SAPSO-LSSVM[J]. Process Automation Instrumentation, 2016, 37(7):88-92.
[21]朱万进，岳峻峰.基于GA-LSSVM直吹式制粉系统煤粉细度软测量[J].上海电力大学学报，2023,39(5):472-478.ZHU Wanjin, YUE Junfeng. Soft measurement of pulverized coal fineness in direct-firing pulverizing system based on GA-LSSVM[J]. Journal of Shanghai University of Electric Power, 2023, 39(5):472-478.
[22]翁文霖，孙玲飞，赵长祥，等.基于信号软测量的再热蒸汽温度控制研究[J].东北电力技术，2024,45(2):1-4.WENG Wenlin, SUN Lingfei, ZHAO Changxiang, et al. Research on reheated steam temperature control based on signal soft sensing[J]. Northeast Electric Power Technology, 2024, 45(2):1-4.
[23]LI K, LI Q Y , FAN H D, et al. Soft sensing of SO2 emission for ultra-low emission coal-fired power plant with dynamic model and segmentation model[J]. Fuel, 2023, 332:125921.
[24]米翠丽，郭江龙，李晓光，等.基于烟气测试的电站锅炉煤质软测量模型[J].热力发电，2015,44(7):62-65+77.MI Cuili, GUO Jianglong, LI Xiaoguang, et al. Soft sensor model for coal quality in power station boilers based on flue gas testing[J]. Thermal Power Generation, 2015, 44(7):62-65+77.
[25]李海永，杨濮亦.火电厂入炉煤低位热值实时测量的研究及其应用[J].电力建设，2013,34(10):60-64.LI Haiyong, YANG Puyi. Research and application of real-time measurement of lower heating value of coal in power plant boilers[J]. Electric Power Construction, 2013, 34(10):60-64.
[26]孟宪鹏.火电厂热工参数软测量技术的应用研究[J].自动化应用，2023,64(19):164-166+172.MENG Xianpeng. Application research of soft sensing technology for thermal parameters in thermal power plants[J]. Automation Application, 2023, 64(19):164-166+172.
[27]邓蓬月.基于深度神经网络的火电厂排放物浓度软测量和预测研究与实现[D].西安：西安电子科技大学，2022.DENG Pengyue. Research and implementation of soft measurement and prediction for emission concentration in thermal power plants based on deep neural network[D]. Xi'an:Xidian University, 2022.
[28]翁卫国，刘博文，郭一杉，等.基于入口NOx浓度软测量的脱硝系统先进控制研究[J].锅炉制造，2019(5):17-22.WENG Weiguo, LIU Bowen, GUO Yishan, et al. Advanced control of denitration system based on soft measurement of inlet NOx concentration[J]. Boiler Manufacturing, 2019(5):17-22.
[29]刘宗奇.SCR脱硝系统NOx浓度软测量及预测控制研究[D].吉林:东北电力大学，2024.LIU Zongqi. Research on soft measurement and predictive control of NOx concentration in SCR denitration system[D]. Jilin:Northeast Electric Power University, 2024.
[30]林伟俊.基于数据驱动的锅炉燃烧动态过程建模与磨煤机运行优化[D].杭州:浙江大学，2023.LIN Weijun. Data-driven modeling of boiler combustion dynamic process and optimization of coal mill operation[D]. Hangzhou:Zhejiang University, 2023.
[31]ADAMS D, OH D H, KIM D W, et al. Prediction of SOx-NOx emission from a coal-fired CFB power plant with machine learning:plant data learned by deep neural network and least square support vector machine[J]. Journal of Cleaner Production,2020, 270:122310.
[32]DING X, FENG C, YU P, et al. Gradient boosting decision tree in the prediction of NOx emission of waste incineration[J]. Energy,2023, 264:126174.
[33]杨红.脱硫系统浆液pH值软测量方法研究[D].北京：华北电力大学，2021.YANG Hong. Research on soft measurement method of slurry pH value in desulfurization system[D]. Beijing:North China Electric Power University, 2021.
[34]梁伟平，薛文雅，马靖宁，等.基于WOA-Elman神经网络的城市固废焚烧炉主蒸汽流量软测量[J/OL].控制工程，1-7[2025-01-09]. https://doi.org/10.14107/j.cnki.kzgc.20220449.LIANG Weiping, XUE Wenya, MA Jingning, et al. Soft measurement of main steam flow in municipal solid waste incinerators based on WOA-Elman neural network[J/OL]. Control Engineering, 1-7[2025-01-09]. https://doi.org/10.14107/j.cnki.kzgc.20220449.
[35]邓怀勇，马琴，陈国彬，等.基于AWOA算法与LSSVM的主蒸汽流量软测量模型[J].仪表技术与传感器，2018(12):78-82.DENG Huaiyong, MA Qin, CHEN Guobin, et al. Soft sensor model for main steam flow based on AWOA algorithm and LSSVM[J]. Instrument Technique and Sensor, 2018(12):78-82.
[36]张恒，刘吉臻，钟亮民，等.一种超超临界直流锅炉水煤比软测量方法[J].电站系统工程，2014,30(5):19-22.ZHANG Heng, LIU Jizhen, ZHONG Liangmin, et al. A soft measurement method for water-coal ratio of ultra-supercritical oncethrough boiler[J]. Power System Engineering, 2014, 30(5):19-22.
[37]张维，刘吉臻.超超临界燃煤机组直流锅炉水煤比软测量方法研究[J].动力工程学报，2016,36(2):99-106.ZHANG Wei, LIU Jizhen. Research on soft measurement method of water-coal ratio for ultra-supercritical coal-fired oncethrough boiler[J]. Journal of Chinese Society of Power Engineering, 2016, 36(2):99-106.
[38]徐红伟，李美双，李崇晟，等.基于数据驱动的汽轮机主蒸汽流量软测量[J].动力工程学报，2023,43(6):742-748.XU Hongwei, LI Meishuang, LI Chongsheng, et al. Data-driven soft sensor for main steam flow of steam turbine[J]. Journal of Chinese Society of Power Engineering, 2023, 43(6):742-748.
[39]李晨光，李希鑫，杨淦，等.催化余热锅炉智能水质分析系统的开发与应用[J].工业水处理，2024, 44(9):181-187.LI Chenguang, LI Xixin, YANG Gan, et al. Development and application of intelligent water quality analysis system for catalytic waste heat boiler[J]. Industrial Water Treatment, 2024,44(9):181-187.
基本信息:
DOI：10.19944/j.eptep.1674-8069.2026.01.004
中图分类号:TM621
引用信息:
[1]王若旭,陈晴,武文斌,等.软测量技术在火电机组中的典型应用研究进展[J].电力科技与环保,2026,42(01):34-43.DOI:10.19944/j.eptep.1674-8069.2026.01.004.
基金信息:
中国南方电网有限责任公司科技项目（NYJS2020KJ005）