|本期目录/Table of Contents|

[1]崔智明.基于物联网的磨煤机风量智能软测量系统设计[J].工业仪表与自动化装置,2023,(06):125-132.[doi:DOI:10.19950/j.cnki.cn61-1121/th.2023.06.022]
　CUI Zhiming.Design of an IoT-based intelligent soft measurement system for coal mill primary air volume[J].Industrial Instrumentation & Automation,2023,(06):125-132.[doi:DOI:10.19950/j.cnki.cn61-1121/th.2023.06.022]
点击复制

基于物联网的磨煤机风量智能软测量系统设计

《工业仪表与自动化装置》[ISSN:1000-0682/CN:61-1121/TH]

卷:
期数:
2023年06期
页码:
125-132
栏目:
出版日期:
2023-12-15

文章信息/Info

Title:
Design of an IoT-based intelligent soft measurement system for coal mill primary air volume
文章编号:
1000-0682（2023）06-0125-00
作者:
崔智明
潞安化工集团能源事业部，山西 长治 046299
Author(s):
CUI Zhiming
LU’ AN CHEMICAL GROUP CO.,LTD. Shanxi Changzhi 046299,China
关键词:
磨煤机一次风量;  软测量;  长短期记忆网络;  PyQt5
Keywords:
primary air volume of coal mill;  Soft measurement;  long short-term memory;  PyQt5
分类号:
TP23
DOI:
DOI:10.19950/j.cnki.cn61-1121/th.2023.06.022
文献标志码:
A
摘要:
磨煤机入口一次风量作为锅炉燃料输送的主要动力，对锅炉燃烧具有主导作用。针对一次风量传统测量难度大且成本高，易受测量仪表和临场环境影响等问题，该文设计了基于物联网(Internet of Things, IoT)的磨煤机风量智能软测量系统。以陕西省榆林市某火电厂为例，首先，建立磨煤机入口一次风量的传感器数据采集系统，通过Lora无线传输至应用层，对上传数据进行预处理，构建模型训练集；其次，建立长短期记忆(Long Short-Term Memory, LSTM)网络软测量初步模型，然后利用PyQt5设计磨煤机风量智能测量软件系统，进行离线训练模型参数调整和在线磨煤机风量测量，并支持数据可视化；最后，设计性能测试实验，与经典反向传播(Back Propagation, BP)模型测量结果比较，LSTM软测量模型得到数据的平均误差小于0.72%，比BP模型测量结果更罚罡。夯芰Ω俊Ｊ笛楸砻鳎萌聿饬肯低诚灾岣吡巳济旱绯У哪苄裕哂幸欢ǖ氖涤眉壑怠?/html>
Abstract:
The primary driving force for fuel transportation in boiler combustion is the primary air volume at the coal mill inlet. Traditional measurement methods for the primary air volume face challenges such as high difficulty and cost, susceptibility to measurement instruments and on-site environmental influences. To address these issues, this study proposes an Internet of Things (IoT)-based intelligent soft measurement system for coal mill primary air volume. Firstly, taking a thermal power plant in Yulin, Shaanxi Province as an example, a sensor data acquisition system for the coal mill primary air volume is established, and the collected data is wirelessly transmitted to the application layer through Lora. The uploaded data is preprocessed and used to construct the training set. Subsequently, a preliminary soft measurement model based on Long Short-Term Memory (LSTM) neural network is established. A coal mill primary air volume intelligent measurement software system is developed using PyQt5, enabling offline training for model parameter adjustment and online measurement of coal mill primary air volume. The software system also supports data visualization. Finally, performance testing experiments are conducted, and the results are compared with those of the classical Back Propagation (BP) model. The LSTM soft measurement model achieves an average error of less than 0.72% for the data, demonstrating higher accuracy, smaller deviation, and stronger generalization ability compared to the BP model. The experimental results indicate that the proposed soft measurement system significantly improves the energy efficiency of coal-fired power plants and holds practical value.

参考文献/References:

[1]LU X, WEI G, YANG S, et al. Properties analysis and optimization of primary air volume in power station[C]//Electronics, Communications and Control (ICECC), 2011 International Conference on. IEEE,2011:3848-3851.

[2]LUCCARINI L, PORRA E, SPAGNI A, et al. Soft sensors for control of nitrogen and phosphorus removal from wastewater by neural networks[J].Water Science Technology,2002, 45(4-5):101-107.
[3]WANG Xiaokai, HUA Lin, WANG Xiaoxuan. Soft measurement model of ring’s dimensions for vertical hot ring rolling process using neural networks optimized by genetic algorithm[J]. Journal of Central South University, 2017, 24(1):17-29.
[4]GONZAGA J C B, MELEIRO L A C, KIANG C, et al. ANN-based soft-sensor for real-time process monitoring and control of an industrial polymerization process[J]. Computers and Chemical Engineering,2009, 33(1):43-49.
[5]XU C, DAI F, XU X, et al. GIS-based support vector machine modeling of earthquake-triggered landslide susceptibility in the Jianjiang River watershed, China[J]. it Geomorphology, 2012, 145-146(none):70-80.
[6]WANG Y, LIN B, DONG Y, et al. Mechanism modeling and validation in ultrasonic vibration assisted drilling with variable cross section drilling tool of brittle materials[J].it International Journal of Advanced Manufacturing Technology, 2019:1-13.
[7]GE Z. Supervised latent factor analysis for process data regression modeling and soft sensor application[J]. IEEE Transactions on Control Systems Technology, 2016, 24(3):1004-1011.
[8]HAO X, ZHANG G, MA S. Deep learning[J].International Journal of Semantic Computing, 2016, 10(03):417-439.
[9]LECUN Y, BENGIO Y, HINTON G. Deep learning[J]. Nature, 2015, 521(7553):436-444.
[10]KANEKO H , FUNATSU K .Classification of the degradation of soft sensor models and discussion on adaptive models[J].Aiche Journal, 2013, 59(7):2339-2347.
[11]GERS F A .Learning to forget: continual prediction with LSTM[J].Neural Computation, 1999.
[12]GREFF K, SRIVASTAVA R K, KOUTNIK Jan, et al. LSTM: a search space odyssey[J]. IEEE Transactions on Neural Networks Learning Systems, 2015, 28(10):2222-2232.
[13]Shuting, Dong, Mingming, et al. LSTM based reserve prediction for bank outlets[J]. Journal of Tsinghua University, Natural Science Edition, 2019,24(1):9.
[14]ALTCHE F , FORTELLE A D L .An LSTM network for highway trajectory prediction[J].IEEE, 2017.
[15]KARIM F, MAJUMDAR S, DARABI H, et al. LSTM fully convolutional networks for time series classification[J]. IEEE Access, 2017, 6(99):1662-1669.
[16]SCHMIDHUBER J. Deep learning in neural networks: an overview[J]. Neural Networks, 2015, 61:85-117.
[17]KOLEN J F, KREMER S C. Gradient flow in recurrent nets: the difficulty of learning longterm dependencies[J]. A Field Guide to Dynamical Recurrent Neural Networks, 2001:1-15.
[18]TAN Y H, CHAN C S. Phi-LSTM: a phrase-based hierarchical LSTM model for image captioning[J]. Asian Conference on Computer Vision, 2016.
[19]SOLTAU H, LIAO H, SSK H. Neural speech recognizer: acoustic-to-word LSTM model for large vocabulary speech recognition[J]. Interspeech, 2016, 1566:3707-3711.
[20]BLUCHE T,LOURADOUR J, MESSINA R. Attend and Read: End-to-End Handwritten Paragraph Recognition with MDLSTM Attention[C]//2017 14th IAPR International Conference on Document Analysis and Recognition (ICDAR).IEEE, 2018.
[21]JOZEFOWICZ R, ZAREMBA W, SUTSKEVER I. An empirical exploration of recurrent network architectures[C]International Conference on International Conference on Machine Learning. JMLR.org, 2015,37.
[22]SALAKHUTDINOV N S M .Unsupervised 1eaming of video representations using 1stms[J].[2023-09-13].
[23]LI Jiulong, ZHOU Lingke. Detecting and identifying gross errors based on “rule”[J]. Computer and Modernization, 2012.

相似文献/References:

备注/Memo

备注/Memo:
收稿日期：2023-08-04

第一作者：
崔智明，男，本科，高级工程师，主要从事煤矿供电、工业自动化、智能化，信息化系统集成与研究工作。
更新日期/Last Update: 1900-01-01