2.765

2022影响因子

(CJCR)

  • 中文核心
  • EI
  • 中国科技核心
  • Scopus
  • CSCD
  • 英国科学文摘

留言板

尊敬的读者、作者、审稿人, 关于本刊的投稿、审稿、编辑和出版的任何问题, 您可以本页添加留言。我们将尽快给您答复。谢谢您的支持!

姓名
邮箱
手机号码
标题
留言内容
验证码

城市固体废物焚烧过程炉温的鲁棒加权异构特征集成预测模型

郭京承 严爱军 汤健

郭京承, 严爱军, 汤健. 城市固体废物焚烧过程炉温的鲁棒加权异构特征集成预测模型. 自动化学报, 2024, 50(1): 121−131 doi: 10.16383/j.aas.c230042
引用本文: 郭京承, 严爱军, 汤健. 城市固体废物焚烧过程炉温的鲁棒加权异构特征集成预测模型. 自动化学报, 2024, 50(1): 121−131 doi: 10.16383/j.aas.c230042
Guo Jing-Cheng, Yan Ai-Jun, Tang Jian. Robust weighted heterogeneous feature ensemble prediction model of temperature in municipal solid waste incineration process. Acta Automatica Sinica, 2024, 50(1): 121−131 doi: 10.16383/j.aas.c230042
Citation: Guo Jing-Cheng, Yan Ai-Jun, Tang Jian. Robust weighted heterogeneous feature ensemble prediction model of temperature in municipal solid waste incineration process. Acta Automatica Sinica, 2024, 50(1): 121−131 doi: 10.16383/j.aas.c230042

城市固体废物焚烧过程炉温的鲁棒加权异构特征集成预测模型

doi: 10.16383/j.aas.c230042
基金项目: 国家自然科学基金(62373017, 62073006), 北京市自然科学基金(4212032)资助
详细信息
    作者简介:

    郭京承:北京工业大学信息学部博士研究生. 主要研究方向为复杂过程建模, 智能优化控制方法. E-mail: guojingcheng@ncut.edu.cn

    严爱军:北京工业大学信息学部教授. 主要研究方向为复杂过程建模, 智能优化控制方法. 本文通信作者. E-mail: yanaijun@bjut.edu.cn

    汤健:北京工业大学信息学部教授. 主要研究方向为小样本数据建模, 城市固体废物处理过程智能控制. E-mail: freeflytang@bjut.edu.cn

Robust Weighted Heterogeneous Feature Ensemble Prediction Model of Temperature in Municipal Solid Waste Incineration Process

Funds: Supported by National Natural Science Foundation of China (62373017, 62073006) and Beijing Natural Science Foundation of China (4212032)
More Information
    Author Bio:

    GUO Jing-Cheng Ph.D. candidate at the Faculty of Information Tech-nology, Beijing University of Technology. His research interest covers complex process modeling and intelligent optimization control method

    YAN Ai-Jun Professor at the Fac-ulty of Information Technology, Bei-jing University of Technology. His research interest covers complex process modeling and intelligent optimization control method. Corresponding author of this paper

    TANG Jian Professor at the Faculty of Information Technology, Bei-jing University of Technology. His research interest covers small sample data modeling and intelligent control of municipal solid waste treatment process

  • 摘要: 针对城市固体废物(Municipal solid waste, MSW)焚烧过程, 数据具有异常值和特征变量维度高时, 炉温预测模型的准确性和泛化能力欠缺的挑战性问题, 提出一种鲁棒加权异构特征集成建模方法, 用于建立城市固体废物焚烧过程炉温预测模型. 首先, 依据焚烧过程机理将高维特征变量划分为异构特征集合, 并采用互信息和相关系数综合评估每组异构特征集合的贡献度; 其次, 采用基于混合t分布的鲁棒随机配置网络(Stochastic configuration network, SCN)构建基模型, 同时确定训练样本的惩罚权重; 最后, 设计一种鲁棒加权负相关学习(Negative correlation learning, NCL)策略, 实现基模型的鲁棒同步训练. 使用国内某城市固体废物焚烧厂的炉温历史数据, 对该方法进行测试. 测试结果表明, 该方法建立的炉温预测模型在准确性和泛化能力方面具有优势.
  • 图  1  MSW焚烧过程工艺流程图

    Fig.  1  MSW incineration process flow chart

    图  2  建模策略图

    Fig.  2  Diagram of the modeling strategy

    图  3  炉排上方热电偶分布情况

    Fig.  3  The distribution diagram of thermocouple on the grate

    图  4  每组异构特征集合对炉温的互信息、相关系数和贡献度对比

    Fig.  4  Comparison of mutual information, correlation coefficient, and contribution of each heterogeneous feature set to furnace temperature

    图  5  $L_{max}$取值为10 ~ 50时, 炉温预测模型在验证集上的误差分布

    Fig.  5  Error distribution of the furnace temperature prediction model on the verification set with $L_{max}$ from 10 ~ 50

    图  6  $\mu$取值为0.1 ~ 0.9时, 炉温预测模型在验证集上的误差分布

    Fig.  6  Error distribution of the furnace temperature prediction model on the verification set with $\mu$ from 0.1 ~ 0.9

    图  7  在不同异常值比例下, 各集成模型对炉温预测的性能对比

    Fig.  7  Performance comparison of ensemble models for furnace temperature prediction under different percentages of abnormal value

    图  8  在20%异常值情况下, 各集成模型输出的散点图

    Fig.  8  Scatter diagram of the output of each ensemble model under 20% abnormal value

    表  1  在不同异常值比例下, 各集成炉温预测模型的测试RMSE (均值$\pm$标准差) (℃)

    Table  1  Test RMSE of each ensemble furnace temperature prediction model under different percentages of abnormal value (mean $\pm$ standard deviation) (℃)

    异常值比例 (%) Mt-RSCNE DNNE SCNE MoGL-SCNE BESCN
    0 16.3 $\pm$ 0.06 18.0 $\pm$ 0.24 16.4 $\pm$ 0.06 17.1 $\pm$ 0.12 17.1 $\pm$ 0.13
    10 16.6 $\pm$ 0.06 18.1 $\pm$ 0.23 16.9 $\pm$ 0.11 17.2 $\pm$ 0.14 17.2 $\pm$ 0.20
    20 16.6 $\pm$ 0.06 19.6 $\pm$ 0.26 18.0 $\pm$ 0.10 18.6 $\pm$ 0.12 17.9 $\pm$ 0.22
    下载: 导出CSV

    表  2  在不同异常值比例下, 各集成炉温预测模型的测试MAE (均值$\pm$标准差) (℃)

    Table  2  Test MAE of each ensemble furnace temperature prediction model under different percentages of abnormal value (mean $\pm$ standard deviation) (℃)

    异常值比例 (%) Mt-RSCNE DNNE SCNE MoGL-SCNE BESCN
    0 12.9 $\pm$ 0.06 14.4 $\pm$ 0.23 13.0 $\pm$ 0.07 13.8 $\pm$ 0.12 13.8 $\pm$ 0.13
    10 13.1 $\pm$ 0.06 14.3 $\pm$ 0.22 13.3 $\pm$ 0.10 13.5 $\pm$ 0.12 13.7 $\pm$ 0.15
    20 13.1 $\pm$ 0.05 15.6 $\pm$ 0.21 14.4 $\pm$ 0.09 15.3 $\pm$ 0.13 14.6 $\pm$ 0.19
    下载: 导出CSV

    A1  炉温预测模型过程变量明细

    A1  Process variable details of furnace temperature prediction model

    序号 名称 单位
    1 干燥炉排左内侧速度 %
    2 干燥炉排左外侧速度 %
    3 干燥炉排右内侧速度 %
    4 干燥炉排右外侧速度 %
    5 燃烧炉排1左内侧速度 %
    6 燃烧炉排1左外侧速度 %
    7 燃烧炉排1右内侧速度 %
    8 燃烧炉排1右外侧速度 %
    9 燃烧炉排2左内侧速度 %
    10 燃烧炉排2左外侧速度 %
    11 燃烧炉排2右内侧速度 %
    12 燃烧炉排2右外侧速度 %
    13 干燥炉排左1空气流量 ${\rm {km^3N/h} }$
    14 干燥炉排右1空气流量 ${\rm {km^3N/h} }$
    15 干燥炉排左2空气流量 ${\rm {km^3N/h} }$
    16 干燥炉排右2空气流量 ${\rm {km^3N/h} }$
    17 燃烧段炉排左1-1段空气流量 ${\rm {km^3N/h} }$
    18 燃烧段炉排右1-1段空气流量 ${\rm {km^3N/h} }$
    19 燃烧段炉排左1-2段空气流量 ${\rm {km^3N/h} }$
    20 燃烧段炉排右1-2段空气流量 ${\rm {km^3N/h} }$
    21 燃烧段炉排左2-1段空气流量 ${\rm {km^3N/h} }$
    22 燃烧段炉排右2-1段空气流量 ${\rm {km^3N/h} }$
    23 燃烧段炉排左2-2段空气流量 ${\rm {km^3N/h} }$
    24 燃烧段炉排右2-2段空气流量 ${\rm {km^3N/h} }$
    25 二次风流量 ${\rm {km^3N/h} }$
    26 一次风机出口空气压力 kPa
    27 干燥段炉排左内侧温度
    28 干燥段炉排左外侧温度
    29 干燥段炉排右内侧温度
    30 干燥段炉排右外侧温度
    31 燃烧段炉排1左内侧温度
    32 燃烧段炉排1左外侧温度
    33 燃烧段炉排1右内侧温度
    34 燃烧段炉排1右外侧温度
    35 燃烧段炉排2左内侧温度
    36 燃烧段炉排2左外侧温度
    37 燃烧段炉排2右内侧温度
    38 燃烧段炉排2右外侧温度
    39 干燥段炉排进口空气温度
    40 燃烧段炉排进口空气温度
    41 一次风加热器出口空气温度
    42 炉温(当前值)
    43 炉温(预测值)
    下载: 导出CSV
  • [1] Zhao X G, Jiang G W, Li A, Li Y. Technology, cost, a performance of waste-to-energy incineration industry in China. Renewable and Sustainable Energy Reviews, 2016, 55(3): 115-130
    [2] Cheng H F, Hu Y A. Municipal solid waste (MSW) as a renewable source of energy: Current and future practices in China. Bioresource Technology, 2010, 101(11): 3816-3824 doi: 10.1016/j.biortech.2010.01.040
    [3] Nzihou A, Themelis N J, Kemiha M, Benhamou Y. Dioxin emissions from municipal solid waste incinerators (MSWIs) in France. Waste Management, 2012, 32(12): 2273-2277 doi: 10.1016/j.wasman.2012.06.016
    [4] Ding H X, Tang J, Qiao J F. Dynamic modeling of multi-input and multi-output controlled object for municipal solid waste incineration process. Applied Energy, 2023, 339(1): Article No. 120982
    [5] Alobaid F, Al-Maliki W A K, Lanz T, Haaf M, Brachthauser A, Epple B, et al. Dynamic simulation of a municipal solid waste incinerator. Energy, 2018, 149(4): 230-249
    [6] He H J, Meng X, Tang J, Qiao J F. A novel self-organizing TS fuzzy neural network for furnace temperature prediction in MSWI process. Neural computing and applications, 2022, 34(12): 9759-9776
    [7] 丁海旭, 汤健, 夏恒, 乔俊飞. 基于TS-FNN的城市固废焚烧过程MIMO被控对象建模. 控制理论与应用, 2022, 39(8): 1529-1540 doi: 10.7641/CTA.2022.10524

    Ding Hai-Xu, Tang Jian, Xia Heng, Qiao Jun-Fei. Modeling of MIMO controlled object in municipal solid waste incineration process based on TS-FNN. Control Theory & Applications, 2022, 39(8): 1529-1540 doi: 10.7641/CTA.2022.10524
    [8] Scardapane S, Wang D H. Randomness in neural networks: an overview. Wiley Interdisciplinary Reviews: Data Mining and Knowledge Discovery, 2017, 7(2): 1-18
    [9] Wang D H, Li M. Stochastic configuration networks: fundamentals and algorithms. IEEE Transactions on Cybernetics, 2017, 47(10): 3346-3479
    [10] Lu J, Ding J L. A novel stochastic configuration network with iterative learning using privileged information and its application. Information Sciences, 2022, 613(10): 953-965
    [11] Wang Q, Hong Q, Wu S, Dai W. Multitarget stochastic configuration network and applications. IEEE Transactions on Artificial Intelligence, 2023, 4(2): 338-348 doi: 10.1109/TAI.2022.3162570
    [12] 代伟, 李德鹏, 杨春雨, 马小平. 一种随机配置网络的模型与数据混合并行学习方法. 自动化学报, 2021, 47(10): 2427-2437 doi: 10.16383/j.aas.c190411

    Dai Wei, Li Da-Peng, Yang Chun-Yu, Ma Xiao-Ping. A model and data hybrid parallel learning method for stochastic configuration networks. Acta Automatica Sinica, 2021, 47(10): 2427-2437 doi: 10.16383/j.aas.c190411
    [13] Wang D H, Li M. Robust stochastic configuration networks with kernel density estimation for uncertain data regression. Information Sciences, 2017, 412(10): 210-222
    [14] Li M, Huang C Q, Wang D H. Robust stochastic configuration networks with maximum correntropy criterion for uncertain data regression. Information Sciences, 2019, 473(4): 73-86
    [15] Lu J, Ding J L. Mixed-distribution based robust stochastic configuration networks for prediction interval construction. IEEE Transactions on Industrial Informatics, 2020, 16(8): 5099-5109 doi: 10.1109/TII.2019.2954351
    [16] Yan A J, Guo J C, Wang D H. Robust stochastic configuration networks based on Student’s-t mixture distribution. Information Sciences, 2022, 607(8): 493-505
    [17] Zaman E A K, Mohamed A, Ahmad A. Feature selection for online streaming high-dimensional data: A state-of-the-art review. Applied Soft Computing, 2022, 127(9): Article No. 109355
    [18] Chandrashekar G, Sahin F. A survey on feature selection methods. Computers and Electrical Engineering, 2014, 40(12): 16-28
    [19] Wang D H, Cui C H. Stochastic configuration networks ensemble with heterogeneous features for large-scale data analytics. Information Sciences, 2017, 417(31): 55-71
    [20] Liu Y, Yao X. Simultaneous training of negatively correlated neural networks in an ensemble. IEEE Transactions On Systems, Man, and Cybernetics—Part B: Cybernetics, 1999, 29(6): 716-725 doi: 10.1109/3477.809027
    [21] Zhuang J B, Tang J, Aljerf L. Comprehensive review on mechanism analysis and numerical simulation of municipal solid waste incineration process based on mechanical grate. Fuel, 2022, 320(4): 1-20
    [22] Xia Z, Shan P, Chen C, Du H, Huang J, Bai L. A two-fluid model simulation of an industrial moving grate waste incinerator. Waste Management, 2020, 104(1): 183-191
    [23] Alhamdoosh M, Wang D H. Fast decorrelated neural network ensembles with random weightInformation Sciences, 2014, 264(11): 104-117
    [24] Lu J, Ding J L, Dai X, Chai T Y. Ensemble stochastic configuration networks for estimating prediction intervals: A simultaneous robust training algorithm and its application.IEEE Transactions on Neural Networks and Learning Systems, 2020, 31(12): 5426-5440 doi: 10.1109/TNNLS.2020.2967816
  • 加载中
图(8) / 表(3)
计量
  • 文章访问数:  252
  • HTML全文浏览量:  111
  • PDF下载量:  63
  • 被引次数: 0
出版历程
  • 收稿日期:  2023-02-09
  • 录用日期:  2023-04-13
  • 网络出版日期:  2023-10-12
  • 刊出日期:  2024-01-20

目录

    /

    返回文章
    返回