Dynamic prediction of pollutants emission from circulating fluidized bed unit based on CNN-GRU-MHA

王勇权; 高明明; 王唯铧; 张鹏新; 成永强

doi:10.19666/j.rlfd.202408194

English

您当前的位置：

首页 >

文章列表页 >

Dynamic prediction of pollutants emission from circulating fluidized bed unit based on CNN-GRU-MHA

Thermal Power Generation Vol. 54, Issue 7, Pages: 33-42(2025)

更新时间：2026-08-11

- Dynamic prediction of pollutants emission from circulating fluidized bed unit based on CNN-GRU-MHA
- 作者机构：
  
  1. 新能源电力系统全国重点实验室(华北电力大学)
  2. 控制与计算机工程学院(华北电力大学)
- 作者简介：
- 基金信息：
- DOI：10.19666/j.rlfd.202408194
  CLC： X773
- Published：2025
- 稿件说明：
移动端阅览
[1]王勇权,高明明,王唯铧,等.基于CNN-GRU-MHA的CFB机组污染物排放动态预测[J],2025,54(07):33-42.
[1]王勇权,高明明,王唯铧,等.基于CNN-GRU-MHA的CFB机组污染物排放动态预测[J],2025,54(07):33-42. DOI： 10.19666/j.rlfd.202408194.

DOI：10.19666/j.rlfd.202408194.

摘要

SO2与NOx排放质量浓度的精准预测可以有效指导污染物排放控制，对CFB机组环保运行具有重要意义。以某330 MW CFB机组为研究对象，采用Pearson相关系数实现输入变量筛选，应用四分位距(interquartile range

IQR)方法筛选并替换异常值，同时进行归一化，完成数据预处理；随后，通过卷积神经网络(convolutional neural network

CNN)提取输入变量的特征，进入门循环单元(gated recurrent unit

GRU)处理时间序列特征，并引入多头自注意力(multi-head attention

MHA)机制捕捉特征之间的重要关系，经训练后反归一化得到模型输出；最后，使用平均绝对误差MAE、平均绝对百分比误差MAPE和决定系数R2评估测试集的结果。结果表明，该CNN-GRU-MHA模型能够较为准确地预测CFB机组的污染物排放质量浓度。消融实验与模型对比证明了该模型的优越性能。该CNN-GRU-MHA模型可以实现CFB机组污染物排放质量浓度的监测与优化指导，从而使电厂及时调整运行参数，确保污染物排放达标。

Abstract

The accurate prediction of SO2 and NOx emission mass concentrations can effectively guide the control of pollutants emissions

which is of great significance for the environmental protection operation of circulating fluidized bed(CFB) units. A 330 MW CFB unit is taken as the research object

and the Pearson coefficient is used to realize the screening of input variables

and the interquartile range(IQR) method is applied to screen the outliers and replace them with the normalization at the same time

to complete the data preprocessing. Subsequently

the features of input variables are extracted by convolutional neural network(CNN)

and by entering into the gate-recurrent unit(GRU) the time-series features are processed. The multi-head self-attention(MHA) mechanism is introduced to capture the important relationships between features

and the model output is obtained after training. Finally

the results of the test set are evaluated using the mean absolute error(MAE)

mean absolute percentage error(MAPE)

and the coefficient of determination(R2). The results show that the model is able to predict the pollutants mass concentration in CFBs more accurately and achieve good prediction results

and the superior performance of the model is proved by the comparison of ablation experiments with the model. The proposed CNN-GRU-MHA model can realize the monitoring and optimization guidance of pollutants emissions CFB units

so that the power plant can adjust the operation parameters in time to ensure that the pollutants emissions meet the standards.

关键词

Keywords

references

姚禹歌,黄中,张缦,等.中国CFB燃烧技术的发展与展望[J].热力发电, 2021, 50(11):13-19.YAO Yuge, HUANG Zhong, ZHANG Man, et al.Development and prospects of circulating fluidized bed combustion technology in China[J]. Thermal Power Generation, 2021, 50(11):13-19.

《全面实施燃煤电厂超低排放和节能改造工作方案》[J].节能与环保, 2016(1):32.Work program for the comprehensive implementation of ultra-low emission and energy-saving retrofit of coal-fired power plants[J]. Energy Conservation&Environmental Protection, 2016(1):32.

张洪福. CFB机组负荷响应特性与灵活运行控制研究[D].北京:华北电力大学, 2023:1.ZHANG Hongfu. Study on load response characteristics and flexible operation control of circulating fluidized bed unit[D]. Beijing:North China Electric Power University,2023:1.

高明明,郭炯楠,于浩洋,等.适应深度调峰的CFBNOx排放建模[J].洁净煤技术, 2023, 29(6):24-31.GAO Mingming, GUO Jiongnan, YU Haoyang, et al.Modeling NOx emissions from circulating fluidized beds adapted to deep peaking[J]. Clean Coal Technology,2023, 29(6):24-31.

杨敬池,王菁,王鹏程,等.超临界350 MW循环流化床锅炉宽负荷NOx排放特性及控制技术[J].洁净煤技术, 2025, 31(1):78-85.YANG Jingchi, WANG Jing, WANG Pengcheng, et al.Characterization and mitigation of NOx emissions across a wide load range in supercritical 350 MW circulating fluidized bed boiler[J]. Clean Coal Technology, 2025,31(1):78-85.

YUE G, CAI R, LU J, et al. From a CFB reactor to a CFB boiler–The review of R&D progress of CFB coal combustion technology in China[J]. Powder Technology,2017, 316:18-28.

KACHAVIMATH A V, NAZARE S V, AKKI S S.Distributed denial of service attack detection using na?ve bayes and k-nearest neighbor for network forensics[C]//Proceedings of the 2020 2nd International Conference on Innovative Mechanisms for Industry Applications(ICIMIA), F, 2020. IEEE.

DONG S. Multi class SVM algorithm with active learning for network traffic classification[J]. Expert Systems with Applications, 2021, 176:114885.

ELMRABIT N, ZHOU F, LI F, et al. Evaluation of machine learning algorithms for anomaly detection[C]//Proceedings of the 2020 International Conference on Cyber Security and Protection of Digital Services(Cyber Security), F, 2020. IEEE.

张砺彦,麦永强,张月辉,等.基于神经网络的大型燃煤电厂SO2和NOx的污染物预测模型[J].热力发电,2008, 37(10):9-13.ZHANG Liyan, MAI Yongqiang, ZHANG Yuehui, et al.Neural network-based pollutant prediction model for SO2and NOx in large coal-fired power plants[J]. Thermal Power Generation, 2008, 37(10):9-13.

LIU L, WANG P, LIN J, et al. Intrusion detection of imbalanced network traffic based on machine learning and deep learning[J]. IEEE Access, 2020, 9:7550-7563.

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.

夏炫昊. CFB机组污染物模型研究与优化[D].北京:华北电力大学, 2023:1.XIA Xuanhao. Pollutant modeling and optimization of circulating fluidized bed units[D]. Beijing:North China Electric Power University, 2023:1.

YU H, GAO M, ZHANG H, et al. Data-driven optimization of pollutant emission and operational efficiency for circulating fluidized bed unit[J]. Energy,2023, 281:128338.

CHEN J, GAO M, ZHANG H, et al. Dynamic prediction of SO2 emission based on hybrid modeling method for coal-fired circulating fluidized bed[J]. Fuel, 2023, 346:128284.

HAN Z, XIE Y, HOSSAIN M M, et al. A hybrid deep neural network model for NOx emission prediction of heavy oil-fired boiler flames[J]. Fuel, 2023, 333:126419.

WANG X, LIU W, WANG Y, et al. A hybrid NOx emission prediction model based on CEEMDAN and AM-LSTM[J]. Fuel, 2022, 310:122486.

SHIN H C, ROTH H R, GAO M, et al. Deep convolutional neural networks for computer-aided detection:CNN architectures, dataset characteristics and transfer learning[J]. IEEE Transactions on Medical Imaging, 2016, 35(5):1285-1298.

张浩,刘振娟,李宏光,等.基于关联变量时滞分析卷积神经网络的生产过程时间序列预测方法[J].化工学报, 2017, 68(9):3501-3510.ZHANG Hao, LIU Zhenjuan, LI Hongguang. Process time series prediction based on application of correlated process variables to CNN time delayed analyses[J].CIESC Journal, 2017, 68(9):3501-3510.

GRAVES A, MOHAMED A R, HINTON G. Speech recognition with deep recurrent neural networks[C]//Proceedings of the 2013 IEEE International Conference on Acoustics, Speech and Signal Processing, F, 2013.IEEE.

SAJJAD M, KHAN Z A, ULLAH A, et al. A novel CNN-GRU-based hybrid approach for short-term residential load forecasting[J]. IEEE Access, 2020, 8:143759-143768.

韩玉民,郝晓燕.基于子词嵌入和相对注意力的材料实体识别[J].计算机应用, 2022, 42(6):1862-1868.HAN Yumin, HAO Xiaoyan. Material entity recognition based on subword embedding and relative attention[J].Journal of Computer Applications, 2022, 42(6):1862-1868.

COHEN I, HUANG Y, CHEN J, et al. Pearson correlation coefficient[J]. Noise Reduction in Speech Processing, 2009, 18:1-4.

王雪松,谷阳阳,程玉虎.基于复杂网络的时延基因调控网络构建[J].电子学报, 2010, 38(11):2518-2522.WANG Xuesong, GU Yangyang, CHENG Yuhu.Construction of delay gene regulatory network based on complex network[J]. Acta Electronica Sinica, 2010,38(11):2518-2522.

WHALEY III D L. The interquartile range:theory and estimation[D]. East Tennessee State University, 2005:1.

郭喜峰,王凯泽,单丹,等.多角度基于CEEMDANCNN-BiLSTM模型的锂离子电池RUL预测[J].太阳能学报, 2024, 45(7):181-189.GUO Xifeng, WANG Kaize, SHAN Dan, et al. RUL prediction for lithium ion batteries based on CEEMDAN-CNN-BiLSTM model from multiple perspectives[J]. Acta Energiae Solaris Sinica, 2024,45(7):181-189.

Views

796

下载量

CSCD

Alert me when the article has been cited

提交

Publicity Resources

基于物理信息神经网络的超临界燃煤发电机炉协调系统动态建模研究

燃煤机组SCR脱硝入口NOx质量浓度的多模型动态预测及协同部署

Application of deep learning algorithm in intelligent operation and maintenance of UAVs in photovoltaic power plants

Application of machine learning in state analysis of power plant equipment

Operation optimization of CFB boiler based on data-driven and pattern mining

Related Author

吴啸

陈凌霄

赵玉林

李留斌

李彦柳

吴健

沈波

卢得勇

Related Institution

东南大学大型发电装备安全运行与智能测控国家工程研究中心

国网江苏省电力有限公司

国网江苏省电力有限公司泰州供电分公司

浙江浙能科技环保集团股份有限公司

浙江浙能兰溪发电有限责任公司

⁰