|本期目录/Table of Contents|

[1]仉元梦,张福群,王鑫.传感器阵列结合BO-CNN的可燃有毒气体安全检测方法研究*[J].中国安全生产科学技术,2025,21(7):120-126.[doi:10.11731/j.issn.1673-193x.2025.07.016]
　ZHANG Yuanmeng,ZHANG Fuqun,WANG Xin.Research on the safety detection method of combustible and toxic gases using sensor array combined with BO-CNN[J].Journal of Safety Science and Technology,2025,21(7):120-126.[doi:10.11731/j.issn.1673-193x.2025.07.016]
点击复制

传感器阵列结合BO-CNN的可燃有毒气体安全检测方法研究*()

《中国安全生产科学技术》[ISSN:1673-193X/CN:11-5335/TB]

卷:
21
期数:
2025年7期
页码:
120-126
栏目:
职业安全卫生管理与技术
出版日期:
2025-07-30

文章信息/Info

Title:
Research on the safety detection method of combustible and toxic gases using sensor array combined with BO-CNN
文章编号:
1673-193X(2025)-07-0120-07
作者:
仉元梦; 张福群; 王鑫
（沈阳化工大学 环境与安全工程学院，辽宁 沈阳 110142）
Author(s):
ZHANG Yuanmeng;  ZHANG Fuqun;  WANG Xin
（College of Environment and Safety Engineering,Shenyang University of Chemical Technology,Shenyang Liaoning 110142,China）
关键词:
可燃有毒气体; 气体分类识别; 检测技术; 传感器阵列
Keywords:
combustible and toxic gases;  gas classification and identification;  detection technology;  sensor array
分类号:
X937
DOI:
10.11731/j.issn.1673-193x.2025.07.016
文献标志码:
A
摘要:
为准确识别化工厂内存在的可燃有毒气体，保障化工作业中生产人员的安全，建立1种硬件和软件相结合的检测系统。首先选取多类型半导体气体传感器探头组成阵列，在实验室模拟环境中对化工过程常见可燃有毒气体进行检测，然后整合传感器阵列输出的数据，提出1种基于贝叶斯优化卷积神经网络的气体识别方法，并验证该模型在气体检测中的有效性。研究结果表明：经过贝叶斯优化的CNN模型在气体的定性识别中，准确率达到99.26%，比传统的CNN模型和BPNN模型的准确率高，且收敛速度更快，训练过程更稳定；构建的传感器阵列和算法模型相结合的检测系统能准确高效地识别可燃有毒气体，适用于化工过程的气体检测。研究结果可为化工安全领域的气体检测识别提供新思路。
Abstract:
In order to accurately identify the combustible and toxic gases in chemical plants and ensure the safety of production personnel during chemical operations,a hardware-software integrated detection system was established.First,multiple semiconductor gas sensor probes are selected to form an array for detecting common combustible and toxic gases in simulated laboratory environments.The collected sensor array data is then integrated to establish a Bayesian-optimized Convolutional Neural Network (BO-CNN) based gas identification method,with experimental validation confirming the model’s effectiveness in gas detection.Research results demonstrate that the Bayesian-optimized CNN model achieves a 99.26% accuracy rate in qualitative gas identification,outperforming traditional CNN and BPNN models in both accuracy and convergence speed while exhibiting greater training stability.The integrated detection system combining sensor arrays with this algorithmic model enables precise and efficient identification of combustible and toxic gases,proving suitable for gas detection in chemical processes.These findings provide innovative approaches to gas detection and identification in chemical process safety.

参考文献/References:

［1］XU Y,ZHAO X,CHEN Y,et al.Research on a mixed gas classification algorithm based on extreme random tree［J］.Applied Sciences,2019,9(9):1728.
［2］SRINIVASULU K,GOVIND S S.Discrimination of gases with a single chemiresistive multi-gas sensor using temperature sweeping and machine learning［J］.Sensors and Actuators B:Chemical,2021，348：130725.
［3］CHU J,LI W,YANG X,et al.Identification of gas mixtures via sensor array combining with neural networks［J］.Sensors and Actuators B:Chemical,2021,329:129090.
［4］温志煌.用于智能电子鼻系统的新型混合气体识别算法研究［D］.深圳：深圳大学,2019.
［5］陈家正.基于气体检测的电子鼻性能优化方法研究［D］.重庆：西南大学,2023.
［6］胡纪年,李雨成,李俊桥,等.基于CNN的矿井外因火灾火源定位方法研究［J］.中国安全生产科学技术,2024,20(3):134-140.HU Jinian,LI Yucheng,LI Junqiao,et al.Study on localization method of mine exogenous fire source based on CNN ［J］.Journal of Safety Science and Technology,2024,20(3):134-140.
［7］吴青云,邹亚囡,史雪莹.基于卷积神经网络的电子鼻分类识别［J］.吉林化工学院学报,2022,39(11):38-41.WU Qingyun,ZOU Yanan,SHI Xueying.Electronic nose classification recognition based on convolutional neural network［J］.Journal of Jilin Institute of Chemical Engineering,2022,39(11):38-41.
［8］缪凯.基于贝叶斯优化算法的长短期记忆网络短期负荷预测［D］.保定：华北电力大学,2021.
［9］杨菁.基于贝叶斯优化下的卷积神经网络的波前校正方法研究［D］.武汉：华中科技大学,2021.
［10］任永良,代岳成,高生亮,等.基于贝叶斯优化的CNN-LSTM的油田注水管网压力预测［J］.数学的实践与认识,2024,54(12):160-174.REN Yongliang,DAI Yuecheng,GAO Shengliang,et al.Pressure prediction of oilfield water injection pipeline network based on Bayesian-optimized CNN-LSTM［J］.Mathematics in Practice and Theory,2024,54(12):160-174.
［11］WANG G,JIA R,LIU J,et al.A hybrid wind power forecasting approach based on Bayesian model averaging and ensemble learning［J］.Renewable Energy,2020,145:2426-2434.
［12］GAO J,LIANG J,LU Y,et al.Automatic SWMM parameter calibration method based on the differential evolution and Bayesian optimization algorithm［J］.Water,2023,15(20):3582.
［13］陈开兵,贾红雯,武斌,等.基于LabVIEW的嵌入式电子鼻系统设计［J］.重庆科技学院学报(自然科学版),2020,22(3):88-91.CHEN Kaibing,JIA Hongwen,WU Bin,et al.Design of embedded electronic nose system based on LabVIEW ［J］.Journal of Chongqing University of Science and Technology (Natural Science Edition),2020,22(3):88-91.
［14］刘伟玲,吴龙焦,张思祥,等.基于电子鼻技术的混合气体识别研究［J］.现代电子技术,2020,43(6):57-60.LIU Weiling,WU Longjiao,ZHANG Sixiang,et al.Research on mixed gas recognition based on electronic nose technology ［J］.Modern Electronics Technique,2020,43(6):57-60.
［15］王鑫,张福群,张金萍.基于主成分分析-粒子群优化算法-支持向量机的混合气体分类识别方法研究［J］.职业卫生与应急救援,2024,42(5):633-638,689.WANG Xin,ZHANG Fuqun,ZHANG Jinping.Study on recognition and classification of gas mixtures based on principal component analysis-particle swarm optimization and support vector machine ［J］.Occupational Health and Emergency Rescue,2024,42(5):633-638,689.

相似文献/References:

备注/Memo

备注/Memo:
收稿日期： 2025-03-27
* 基金项目： 辽宁省教育厅自然科学基金项目(LJ2020027)
作者简介： 仉元梦，硕士研究生，主要研究方向为化工安全。
通信作者： 张福群，博士，教授，主要研究方向为化工企业危险源辨识、安全评价与安全控制技术、风险预警及监控。
更新日期/Last Update: 2025-07-28