基于机器学习的双参数火灾探测方法

doi:10.16265/j.cnki.issn1003-3033.2022.05.0874

中国安全科学学报 ›› 2022, Vol. 32 ›› Issue (5): 90-96.doi: 10.16265/j.cnki.issn1003-3033.2022.05.0874

基于机器学习的双参数火灾探测方法

刘全义¹(), 朱博¹, 邓力¹, 石航¹, 梁光华²^,^**()

¹ 中国民用航空飞行学院民航安全工程学院,四川广汉 618307
² 清华大学合肥公共安全研究院,安徽合肥 230601

收稿日期:2021-12-08 修回日期:2022-03-12 出版日期:2022-08-17 发布日期:2022-11-28
通讯作者: 梁光华

作者简介:

刘全义 (1987—),男,河南郸城人,博士,副教授,硕士生导师,主要从事公共安全、航空消防与安全、新一代机载灭火技术等方面的研究。E-mail: quanyiliu2005@126.com。

基金资助:
国家自然科学基金资助(U2033206); 国家自然科学基金资助(U1933105); 四川省科技计划项目(2018GZYZF0069); 四川省科技计划项目(2020YFG0447); 中国民用航空飞行学院基金资助(J2020-110); 中国民用航空飞行学院基金资助(J2020-120); 中国民用航空飞行学院基金资助(X2019-022); 中国民用航空飞行学院大学生双创建设项目(S202010624137)

Double parameters fire detection method based on machine learning

LIU Quanyi¹(), ZHU Bo¹, DENG Li¹, SHI Hang¹, LIANG Guanghua²^,^**()

¹ College of Civil Aviation Safety Engineering, Civil Aviation Flight University of China, Guanghan Sichuan 618307, China
² Hefei Institute for Public Safety Research, Tsinghua University, Hefei Anhui 230601,China

Received:2021-12-08 Revised:2022-03-12 Online:2022-08-17 Published:2022-11-28
Contact: LIANG Guanghua

摘要/Abstract

摘要：

为解决单一技术在火灾探测上造成的误报、漏报现象,设计并建立可燃物燃烧试验平台,选取燃烧产物中质量浓度迅速升高的PM10及CO作为分类算法的特征参数,对特征参数进行数据处理后,采用逻辑回归(LR)、线性判别分析(LDA)、k近邻算法(kNN)、分类与回归树(CART)、朴素贝叶斯与支持向量机(SVM)等6种机器学习算法建立火灾探测模型,并评估分析其分类性能。结果表明: 6种算法中kNN评估准确率、召回率、F₁值和kappa值均高于其他算法,且评估准确率达到95.2%,能够准确地识别燃烧状态;通过分类处理燃烧产物中快速变化的PM10及CO质量浓度,能够较准确识别火灾。

关键词: 机器学习, 火灾探测, PM10, 分类算法, k近邻算法(kNN)

Abstract:

In order to address false alarms or alarm failure caused by a single technology in fire detection, a combustion experimental platform was designed and built. PM10 and CO, concentration of which increased rapidly in the flue gas after fire, were selected as characteristic parameters.Then, data processing was conducted for them, and a fire detection model was established by adopting six machine learning algorithms, including logistic regression(LR), linear discriminant analysis (LDA), kNN algorithm, classification and regression tree(CART), naive Bayes, and support vector machine (SVM). Finally, the model's classification performance was assessed. The results show that among six algorithms, kNN features higher evaluation accuracy, recall rate, F₁ value and kappa value, with its accuracy of assessment reaching as high as 95.2%, making it possible to accurately identify combustion state. This method can accurately detect fire by classifying rapidly changing concentrations of PM10 and CO in combustion products.

Key words: machine learning, fire detection, PM10, classification algorithm, k-nearest neighbor algorithm(kNN)

刘全义, 朱博, 邓力, 石航, 梁光华. 基于机器学习的双参数火灾探测方法[J]. 中国安全科学学报, 2022, 32(5): 90-96.

LIU Quanyi, ZHU Bo, DENG Li, SHI Hang, LIANG Guanghua. Double parameters fire detection method based on machine learning[J]. China Safety Science Journal, 2022, 32(5): 90-96.

图/表 11

图1

图2

图3

图4

图5

图6

图7

图8

表1

图9

图10

参考文献 14

[1]	BU Fengju, GHARAJEH M S. Intelligent and vision-based fire detection systems: a survey[J]. Image and Vision Computing, 2019, 91: 103803.1-103803.15.
[2]	张海云, 王阿川. 基于BP神经网络在森林防火专家系统不确定推理中的应用研究[J]. 中国安全科学学报, 2010, 20(7): 38-42.
	ZHANG Haiyun, WANG Achuan. Study on the application BP neural network in the uncertainty reasoning of forest fire prevention expert system[J]. China Safety Science Journal, 2010, 20(7):38-42.
[3]	何永勃, 张文杰, 杨伟, 等. 飞机货舱复合烟雾探测方法研究[J]. 中国安全科学学报, 2019, 29(1):43-48. doi: 10.16265/j.cnki.issn1003-3033.2019.01.008
	HE Yongbo, ZHANG Wenjie. YANG Wei, et al. Research on multi-sensor smoke detection method for aircraft cargo compartment[J]. China Safety Science Journal, 2019, 29(1):43-48. doi: 10.16265/j.cnki.issn1003-3033.2019.01.008
[4]	文超, 李忠灿, 黄平, 等. 数据驱动的列车晚点传播研究[J]. 中国安全科学学报, 2019, 29(增2):1-9.
	WEN Chao, LI Zhongcan, HUANG Ping, et al. Progress and perspective of data-driven train delay propagation[J]. China Safety Science Journal, 2019, 29(S2): 1-9. doi: 10.16265/j.cnki.issn1003-3033.2019.S2.001
[5]	金杉, 金志刚. 基于自适应模糊广义回归神经网络的区域火灾数据推理预测[J]. 计算机应用, 2015, 35(5):1499-1504. doi: 10.11772/j.issn.1001-9081.2015.05.1499
	JIN Shan, JIN Zhigang. Reasoning and forecasting of regional fire data based on adaptive fuzzy generalized regression neural network[J]. Journal of Computer Applications, 2015, 35(5):1499-1504. doi: 10.11772/j.issn.1001-9081.2015.05.1499
[6]	王志强, 李立君, 黄雁, 等. 基于模糊最小二乘支持向量机的火灾信号辨识[J]. 中南大学学报:自然科学版, 2013, 44(1):202-207.
	WANG Zhiqiang, LI Lijun, HUANG Yan, et al. Fire disaster signal recognition based on fuzzy least squares support vector machines[J]. Journal of Central South University :Science and Technology, 2013, 44(1):202-207.
[7]	EMMY P C, VINSLEY S S, SURESH S. Efficient flame detection based on static and dynamic texture analysis in forest fire detection[J]. Fire Technology, 2018, 54(1):255-288. doi: 10.1007/s10694-017-0683-x
[8]	SAEED F, PAUL A, HONG W H, et al. Machine learning based approach for multimedia surveillance during fire emergencies[J]. Multimedia Tools and Applications, 2020, 79(23):16 201-16 217.
[9]	QIU Xuanbing, WEI Yongbo, LI Ning, et al. Development of an early warning fire detection system based on a laser spectroscopic carbon monoxide sensor using a 32-bit system-on-chip[J]. Infrared Physics & Technology, 2019, 96: 44-51.
[10]	王湛, 朱国庆, 柴国强, 等. 基于机器学习的火灾事故等级分类研究[J]. 消防科学与技术, 2020, 39(12):1735-1739.
	WANG Zhan, ZHU Guoqing, CHAI Guoqiang, et al. Research on fire classification based on machine learning[J]. Fire Science and Technology, 2020, 39(12):1735-1739.
[11]	VASILIEV A A, GRIGORIEVG Y, LAGUTIN A S. Contemporary technologies of early detection of fire in space vehicles[J]. Acta Astronautica, 2017, 135: 76-82. doi: 10.1016/j.actaastro.2016.11.005
[12]	RICHA S, SHALLI R, IMRAN M. A smart approach for fire prediction under uncertain conditions using machine learning[J]. Multimedia Tools and Applications: An International Journal, 2020, 79: 28155-28168.
[13]	VASILIEV A A, GRIGORIEVG Y, LAGUTIN A S. Contemporary technologies of early detection of fire in space vehicles[J]. Acta Astronautica, 2017, 135: 76-82. doi: 10.1016/j.actaastro.2016.11.005
[14]	HJ633-2012, 环境空气质量评价技术规范[S].
	HJ633-2012, Technical specifications for environmental air quality assessment[S].

类别	实际0	实际1
预测0	a	b
预测1	c	d

基于机器学习的双参数火灾探测方法

Double parameters fire detection method based on machine learning

RichHTML

PDF

可视化

摘要/Abstract

引用本文

使用本文

图/表 11

参考文献 14

相关文章 11

编辑推荐

Metrics

本文评价

[1]	张凯, 张科. 基于LightGBM算法的边坡稳定性预测研究[J]. 中国安全科学学报, 2022, 32(7): 113-120.
[2]	刘坤, 焦钰博, 张晓明, 陈晓宇, 蒋朝哲. 基于心电的铁路列车驾驶压力检测研究[J]. 中国安全科学学报, 2022, 32(6): 31-37.
[3]	杨文臣, 周燕宁, 田毕江, 郭凤香, 胡澄宇. 基于聚类分析和SVM的二级公路交通事故严重度预测[J]. 中国安全科学学报, 2022, 32(5): 163-169.
[4]	王斌，施祖建，匡蕾. 基于改进KFDA和RW v-SVM的化工生产系统故障快速诊断[J]. 中国安全科学学报, 2013, 23(8): 84-.
[5]	张昊，赵金勇，李宏文，王燕平. 地铁运营工况点型探测器探测性能试验研究[J]. 中国安全科学学报, 2011, 21(8): 43-.
[6]	邢志祥，陈露. 基于MATLAB的火灾探测报警系统寿命分布和可靠度研究[J]. 中国安全科学学报, 2011, 21(6): 48-.
[7]	张鑫，隋金雪，张岩. 信息融合技术在火灾探测中的应用研究[J]. 中国安全科学学报, 2011, 21(6): 94-.
[8]	王皎，高扬. 基于火灾动力学模拟的古建筑火灾探测器布置研究[J]. 中国安全科学学报, 2011, 21(3): 71-.
[9]	张海云，王阿川. 基于BP神经网络在森林防火专家系统不确定推理中的应用研究[J]. 中国安全科学学报, 2010, 20(7): 38-.
[10]	谢启源，袁宏永，吴龙标，储胜华. 潮湿引起感烟火灾探测器误报研究[J]. 中国安全科学学报, 2004, 14(1): 84-.
[11]	胡毅亭，乔小玲，陈网桦，彭金华，刘荣海. 火炸药企业重大事故隐患危险性评估的计算机自学习实现[J]. 中国安全科学学报, 1999, 9(6): 35-.