|本期目录/Table of Contents|

[1]陈龙龙,李 凌.基于机器学习的湿化仪温度预测建模与仿真[J].工业仪表与自动化装置,2022,(02):56-61+115.[doi:10.19950/j.cnki.cn61-1121/th.2022.02.012]
 CHEN Longlong,LI Ling.Modeling and Simulation of temperature prediction of humidifier based on machine learning[J].Industrial Instrumentation & Automation,2022,(02):56-61+115.[doi:10.19950/j.cnki.cn61-1121/th.2022.02.012]
点击复制

基于机器学习的湿化仪温度预测建模与仿真

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

卷:
期数:
2022年02期
页码:
56-61+115
栏目:
出版日期:
2022-04-15

文章信息/Info

Title:
Modeling and Simulation of temperature prediction of humidifier based on machine learning
文章编号:
1000-0682(2022)02-0000-00
作者:
陈龙龙李 凌
(沈阳化工大学 信息工程学院,辽宁 沈阳 110142)
Author(s):
CHEN Longlong LI Ling
(Shenyang University of Chemical Technology,Liaoning Shengyang 110142,China)
关键词:
湿化仪多元线性回归随机森林支持向量机温度预测
Keywords:
humidifier Multiple linear regression random forest support vector machine temperature prediction
分类号:
TP273
DOI:
10.19950/j.cnki.cn61-1121/th.2022.02.012
文献标志码:
A
摘要:
高流量呼吸湿化仪作为一种典型的非线性、大滞后且对象参数随流量变化的变结构温度控制系统,出气温度会随流量不断变化且易受到各种参数影响。鉴于对象的特殊性及复杂性考虑,无法通过传统的机理建模方法进行模型搭建。针对这一问题的出现,以大量的历史数据为基础,使用机器学习方法进行建模。整个建模过程,选用多元线性回归、随机森林和支持向量机三种算法进行模型训练;在性能效果评估方面,选用可解释方差(VS)、平均绝对误差(MAE)、均方误差(MSE)、判定系数(R2)四种指标进行效果评估。仿真实验结果表明,在不同工况条件下,三种模型皆可以对出气温度做出预测,且通过性能对比分析,相同条件下随机森林的预测效果要优于另外两种算法。实现对出气温度预测,不仅使湿化仪的性能得到提升,同时对于仪器的安全使用也具备一定的指导意义。
Abstract:
As a typical variable structure temperature control system with nonlinear, large delay and object parameters changing with flow, the outlet temperature will change with flow and is easy to be affected by various parameters. Considering the particularity and complexity of the object, the traditional mechanism modeling method can not be used to build the model. To solve this problem, machine learning method is used to model based on a large number of historical data. In the whole modeling process, three algorithms, multiple linear regression, random forest and support vector machine, are selected for model training; In terms of performance effect evaluation, interpretable variance (VS), mean absolute error (MAE), mean square error (MSE) and decision coefficient (R2) are selected for effect evaluation. The simulation results show that the three models can predict the outlet temperature under different working conditions, and through the performance comparative analysis, the prediction effect of random forest under the same conditions is better than the other two algorithms. The prediction of the outlet temperature not only improves the performance of the humidifier, but also has certain guiding significance for the safe use of the instrument.

参考文献/References:

[1]成人经鼻高流量湿化氧疗临床规范应用专家共识[J].中华结核和呼吸杂志,2019(02):83-91.

[2]殷凯生.呼吸衰竭[M].北京:科学出版社,2010:145-146.
[3]葛慧青,代冰.高流量氧疗使用手册[M].沈阳:辽宁科学出版社,2018: 5-55.
[4]詹庆元,李刚.ICU气道湿化精要[M].北京:北京大学医学出版社,2014:132.
[5]魏立峰,李雨桐,杨私,等.多模型内模控制的湿化仪温控系统研[J].电子器件,2020,43(03):667-671.
[6]陈卓,孙龙祥.基于深度学习LSTM网络的短期电力负荷预测方法[J].电子技术,2018,47(01):39-41.
[7]李强,白建荣,李振林.基于 Python 的数据批处理技术探讨及实 现[J].地理空间信息,2015(2):54-56.?/div>
[8]王琦,赵静,胡磊,等.基于机器学习LSTM网络的SO2排放浓度预测[J].自动化与仪表,2021,36(07):77-80+85.
[9]党婷,彭乃驰.基于GM-ES-GASVR组合模型的丽江国内旅游需求预测[J].数学的实践与认识,2017,47(08):279-287.
[10]GEAVES A.Supervised sequence labelling with recurrentneural networks[M].Berlin:Springer,2012.
[11]郭萱.基于随机森林的电影票房预测研究[D].北京:中国石油大学(北京),2018:43-45.
[12]LAHOUAR A, SLAMA J B H. Hour-ahead wind power forecast based on random forests[J]. 0960-1481,2017,Renewableenergy,109:529-541.
[13]MOUNTRAKIS G, IM J, OGOLE C. Support Vector Machines in Remote Sensing: A Review[J]. Isprs Journal Photogrammetry and Remote Sensing, 2011,66(3):247-259.
[14]ZHU G, BLUMBERG D G. Classification Using ASTER Data and SVM Algorithms: the Case Study of Beer Sheva, Israel[J]. Remote Sensing of Environment, 2002, 80(2): 233-240.?div>[15]BALDOMERO-NARANJO M,MARTINEZ-MERI no L I, RODRIGUEZ-CHIA A M. Tightening Big Ms in Integer Programming Formulations for Support Vector Machines with Ramp Loss[J]. European Journal of Operational Research, 2020, 286(1): 84-100.
[16]ZUO W M, WANG F Q, ZHANG D, et al. Distance Metric Learning Via Iterated Support Vector Machines[J]. IEEE Transactions on ImagProcessing, 2017, 26(10): 4937-4950.
[17]刘烁. 乳腺癌数据分析与计算建模[D].福州:福建师范大学,2018:51-53.
[18]朱润苏.基于机器学习的空气质量预测算法研究[D].无锡:江南大学,2021:46-48.
[19]王永乐.基于LSTM的煤炭价格预测模型研究[D].西安:西安科技大学,2020:34-36.

相似文献/References:

[1]董维振,陈 燕*,李媛媛.基于多元逐步回归的带钢性能预测模型[J].工业仪表与自动化装置,2022,(02):107.[doi:10.19950/j.cnki.cn61-1121/th.2022.02.022]
 DONG Weizhen,CHEN Yan*,LI Yuanyuan.Research on prediction model of steel properties based on multiple stepwise regression and data mining[J].Industrial Instrumentation & Automation,2022,(02):107.[doi:10.19950/j.cnki.cn61-1121/th.2022.02.022]

备注/Memo

备注/Memo:
收稿日期:2021-11-09

基金项目:
国家自然科学基金资助项目(61503257)

作者简介:
陈龙龙(1995),男,汉,陕西西安,研究生在读,主要研究方向为智能控制方法研究。

通信作者:
李凌(1972),女,汉,山东梁山,副教授,博士,主要研究方向为复杂系统建模与控制方法研究。
更新日期/Last Update: 1900-01-01