基于机器学习的煤自然发火期预测

doi:10.16265/j.cnki.issn1003-3033.2022.12.0134

中国安全科学学报 ›› 2022, Vol. 32 ›› Issue (12): 118-124.doi: 10.16265/j.cnki.issn1003-3033.2022.12.0134

基于机器学习的煤自然发火期预测

张利冬¹(), 宋泽阳¹^,²^,^**(), 罗振敏¹^,³, 赵珊珊¹

¹ 西安科技大学安全科学与工程学院,陕西西安 710054
² 陕西省煤火灾害防治重点实验室,陕西西安710054
³ 陕西省工业过程安全与应急救援工程技术研究中心,陕西西安 710054

收稿日期:2022-07-15 修回日期:2022-10-13 出版日期:2022-12-28
通讯作者:
^** 宋泽阳(1986—),男,湖南长沙人,博士,副教授,主要从事能源与化工安全、节能环保方面的研究。E-mail:songzeyang03@126.com。
作者简介:
张利冬 (1998—),女,山东临沂人,硕士研究生,研究方向为煤自燃防治技术。E-mail:zld214159@163.com。
基金资助:
国家自然科学基金资助(51804168); 陕西省高层次青年科技人才项目

Prediction of coal spontaneous combustion period based on machine learning

ZHANG Lidong¹(), SONG Zeyang¹^,²^,^**(), LUO Zhenmin¹^,³, ZHAO Shanshan¹

¹ School of Safety Science and Engineering, Xi'an University of Science and Technology, Xi'an Shaanxi 710054,China
² Shaanxi Key Laboratory of Prevention and Control of Coal Fire, Xi'an Shaanxi 710054, China
³ Shaanxi Engineering Research Center for Industrial Process Safety and Emergency Rescue, Xi'an Shaanxi 710054,China

Received:2022-07-15 Revised:2022-10-13 Published:2022-12-28

摘要/Abstract

摘要：

为快速准确地预测煤自然发火期,首先基于大型煤自燃低温氧化试验及文献数据组成数据集,并考虑煤自燃影响因素众多且与发火期存在复杂的非线性关系,建立包含煤自然发火期、环境温度、煤炭热值、水分等参数的数据集;其次采用多层感知机(MLP)和随机森林(RF)等机器学习方法建立煤自然发火期预测模型,表征内部因素和外部因素对发火期的影响;同时为增强模型的拟合能力和泛化能力,利用特征工程研究特征变量的相关性,以筛选模型的输入特征;然后利用网格搜索法优化模型超参数,以提高模型的预测能力;最后利用学习曲线法评估模型状态,防止模型过拟合。结果表明:RF和MLP模型均能预测煤自然发火期,RF模型的泛化能力更高;RF和MLP模型预测的平均绝对误差(MAE)分别为9.34天和12.10天,说明机器学习模型可同时考虑多个内外影响因素的复杂作用。

关键词: 煤自然发火期, 机器学习, 预测模型, 随机森林(RF), 多层感知机(MLP)

Abstract:

In order to predict the spontaneous combustion period of coal quickly and accurately, firstly, based on the large-scale coal spontaneous combustion low-temperature oxidation test and literature data, the data was composed, and considering the many influencing factors of coal spontaneous combustion and the complex nonlinear relationship with the combustion period, the data set including the spontaneous combustion period of coal, ambient temperature, coal calorific value, moisture and other parameters was established. Secondly, machine learning models such as MLP and RF are used to establish a coal spontaneous combustion period prediction model to characterize the influence of internal and external factors on the combustion period. At the same time, in order to enhance the fitting ability and generalization ability of the model, feature engineering is used to study the correlation of feature variable to filter the input features of the model. Then, the gridsearch method is used to optimize the model hyperparameters to improve the prediction ability of the model. Finally, the learning curve method is used to evaluate the model state to prevent overfitting. The results show that both RF and MLP models can predict cola spontaneous combustion period, and the generalization capability of the RF model is higher. The mean absolute error (MAE) predicted by the RF and MLP models were 9.34 days and 12.10 days, respectively, indicating that the machine learning model can consider the complex effects of multiple internal and external factors.

Key words: coal spontaneous combustion period, machine learning, prediction model, random forest (RF), multilayer perceptron (MLP)

张利冬, 宋泽阳, 罗振敏, 赵珊珊. 基于机器学习的煤自然发火期预测[J]. 中国安全科学学报, 2022, 32(12): 118-124.

ZHANG Lidong, SONG Zeyang, LUO Zhenmin, ZHAO Shanshan. Prediction of coal spontaneous combustion period based on machine learning[J]. China Safety Science Journal, 2022, 32(12): 118-124.

图/表 10

图1

图2

图3

图4

图5

图6

图7

表1

图8

图9

参考文献 16

[1]	李林, 姜德义, BEAMISH B B. 基于绝热实验活化能解算煤自然发火期[J]. 煤炭学报, 2010, 35(5):802-805.
	LI Lin, JANG Deyi, BEAMISH B B. Calculation of ignition times under adiabatic conditions by activation energy[J]. Journal of China Coal Society, 2010, 35(5):802-805.
[2]	余明高, 黄之聪, 岳超平. 煤最短自然发火期解算数学模型[J]. 煤炭学报, 2001, 26(5):516-519.
	YU Minggao, HUANG Zhicong, YUE Chaoping. Mathematic model for calculating the shortest coal spontaneous combustion time[J]. Journal of China Coal Society, 2001, 26(5):516-519.
[3]	陈晓坤, 王晨熹, 翟小伟. 基于热爆炸理论的煤燃点确定及动力学分析[J]. 中国安全科学学报, 2022, 32(2):59-65. doi: 10.16265/j.cnki.issn1003-3033.2022.02.009
	CHEN Xiaokun, WANG Chenxi, ZHAI Xiaowei. Determination and kinetics analysis of coal ignition point based on thermal explosion method[J]. China Safety Science Journal, 2022, 32(2):59-65. doi: 10.16265/j.cnki.issn1003-3033.2022.02.009
[4]	张巨峰, 施式亮, 鲁义, 等. 大数据下瓦斯与煤自燃共生灾害智能预警系统:数据特征、应用架构、关键技术[J]. 中国安全科学学报, 2021, 31(9):60-66. doi: 10.16265/j.cnki.issn1003-3033.2021.09.009
	ZHANG Jufeng, SHI Shiliang, LU Yi, et al. Intelligent early warning system of gas and coal spontaneous combustion disaster based on big data:data characteristics,application structure and key technologies[J]. China Safety Science Journal, 2021, 31(9):60-66. doi: 10.16265/j.cnki.issn1003-3033.2021.09.009
[5]	YUAN Han, RICHTER F, REIN G. A multi-step reaction scheme to simulate self-heating ignition of coal:effects of oxygen adsorption and smouldering combustion[J]. Proceedings of the Combustion Institute, 2020, 38:4717-4725. doi: 10.1016/j.proci.2020.07.016
[6]	WU Dejian, SONG Zeyang, SCHMIDT M, et al. Theoretical and numerical study on ignition behaviour of coal dust layers on a hot surface with corrected kinetic parameters[J]. Journal of Hazardous Materials, 2019, 368:156-162. doi: S0304-3894(19)30038-X pmid: 30673621
[7]	JANES A, CARSON D, ACCORSI A, et al. Correlation between self-ignition of a dust layer on a hot surface and in baskets in an oven[J]. Journal of Hazardous Materials, 2008, 159(2/3):528-535. doi: 10.1016/j.jhazmat.2008.02.057
[8]	谭波, 邵壮壮, 郭岩, 等. 基于指标气体关联分析的煤自燃分级预警研究[J]. 中国安全科学学报, 2021, 31(2):33-39. doi: 10.16265/j.cnki.issn 1003-3033.2021.02.005
	TAN Bo, SHAO Zhuangzhuang, GUO Yan et al. Research on grading and early warning of coal spontaneous combustion based on correlation analysis of index gas[J]. China Safety Science Journal, 2021, 31(2):33-39. doi: 10.16265/j.cnki.issn 1003-3033.2021.02.005
[9]	SAHU H B, PADHEE S, MAHAPATRA S S. Prediction of spontaneous heating susceptibility of Indian coals using fuzzy logic and artificial neural network models[J]. Expert Systems with Applications, 2011, 38(3):2271-2282. doi: 10.1016/j.eswa.2010.08.015
[10]	LEI Changkui, DENG Jun, CAO Kai, et al. A random forest approach for predicting coal spontaneous combustion[J]. Fuel, 2018, 223:63-73. doi: 10.1016/j.fuel.2018.03.005
[11]	BI Haobo, WANG Chengxin, LIN Qizhao, et al. Combustion behavior, kinetics, gas emission characteristics and artificial neural network modeling of coal gangue and biomass via TG-FTIR[J]. Energy, 2020, 213:DOI:10.1016/j.energy.2020.118790. doi: 10.1016/j.energy.2020.118790
[12]	张嬿妮, 张帆, 邓军, 等. 不同煤层煤氧化自燃性实验分析[J]. 中国安全生产科学技术, 2018, 14(10):39-45.
	ZHANG Yanni, ZHANG Fan, DENG Jun, et al. Experimental analysis on characteristics of oxidation and spontaneous combustion of coal in different coal seams[J]. Journal of Safety Science and Technology, 2018, 14(10):39-45.
[13]	马砺, 雷昌奎, 张志鹏, 等. 基于GM(0,N)模型的煤自然发火期预测[J]. 辽宁工程技术大学学报:自然科学版, 2016, 35(9):902-907.
	MA Li, LEI Changkui, ZHANG Zhipeng, et al. Forecasting study on spontaneous combustion period of coal based on GM(0,N) model[J]. Journal of Liaoning Technical University:Natural Science, 2016, 35(9):902-907.
[14]	ZHAO Haiquan, LU Lu, HE Zhengyou, et al. Adaptive recursive algorithm with logarithmic transformation for nonlinear system identification in α-stable noise[J]. Digital Signal Processing, 2015, 46:120-132. doi: 10.1016/j.dsp.2015.08.004
[15]	IBRAHIM A I, KHATIB T. A novel hybrid model for hourly global solar radiation prediction using random forests technique and firefly algorithm[J]. Energy Conversion and Management, 2017, 138:413-425. doi: 10.1016/j.enconman.2017.02.006
[16]	BIAN Bing, LIU Zhijian, ZHANG Wei. Studied on indicator gases of coal based on BP neural network forecasting system[J]. Applied Mechanics and Materials, 2014, 685:259-262. doi: 10.4028/www.scientific.net/AMM.685.259

模型	模型性能指标	ξ₁/d	ξ₂/d	ξ₃
RF 模型	训练集	7.56	10.57	0.84
RF 模型	测试集	9.34	12.83	0.67
MLP模型	训练集	7.31	10.18	0.84
MLP模型	测试集	12.10	14.75	0.67

基于机器学习的煤自然发火期预测

Prediction of coal spontaneous combustion period based on machine learning

RichHTML

PDF

可视化

摘要/Abstract

引用本文

使用本文

图/表 10

参考文献 16

相关文章 15

编辑推荐

Metrics

本文评价

[1]	王军武, 何娟娟, 宋盈辉, 刘一鹏, 陈兆, 郭婧怡. 基于RF-SFLA-SVM的装配式建筑高空作业工人不安全行为预警[J]. 中国安全科学学报, 2024, 34(3): 1-8.
[2]	牛甜辉, 耿佃桥, 苑轶, 赵亮, 董辉, 王柏. 基于火灾痕迹的起火点判定研究现状及展望[J]. 中国安全科学学报, 2024, 34(1): 238-246.
[3]	李德龙, 成钰, 李友东. 突发公共卫生事件下地铁安检员需求预测模型[J]. 中国安全科学学报, 2023, 33(S1): 228-234.
[4]	汪伟, 崔欣超, 祁云, 梁然, 贾宝山, 薛凯隆. 基于SSA-RF的采空区煤自燃温度回归分析模型[J]. 中国安全科学学报, 2023, 33(9): 136-141.
[5]	温惠英, 黄坤火, 赵胜. 基于机器学习的高速公路大型货车追尾风险预测[J]. 中国安全科学学报, 2023, 33(9): 173-180.
[6]	靳春玲, 姬照泰, 贡力, 安祥, 周一. 基于WOA-SVM的引水隧洞岩爆烈度评估模型[J]. 中国安全科学学报, 2023, 33(9): 41-48.
[7]	吴方浩, 陈伟, 孙惠中, 牛力, 付建华. 建筑工程安全文明施工费预测精度研究[J]. 中国安全科学学报, 2023, 33(8): 51-58.
[8]	汪伟, 梁然, 祁云, 贾宝山, 武泽伟. 基于PSO-BPNN的煤自燃危险性预测模型[J]. 中国安全科学学报, 2023, 33(7): 127-132.
[9]	段在鹏, 李帆, 郭进, 李炯. 城市易涝区房屋结构安全集成预警模型[J]. 中国安全科学学报, 2023, 33(7): 173-180.
[10]	赵伟, 李书全. 基于机器学习的建筑施工人员安全能力预测模型[J]. 中国安全科学学报, 2023, 33(7): 51-57.
[11]	张西良, 焦灏恺, 李二宝. 基于迁移学习算法的深部爆破振动速度预测[J]. 中国安全科学学报, 2023, 33(6): 64-72.
[12]	段在鹏, 李炯, 李帆, 刘碧强. 农村改造房屋结构安全预警模型[J]. 中国安全科学学报, 2023, 33(4): 100-106.
[13]	何淑波, 项薇, 石钟淼. 基于机器学习的电动汽车电池系统的风险预警[J]. 中国安全科学学报, 2023, 33(2): 159-165.
[14]	李帅, 练章华, 丁亮亮, 廖涛. 多因素耦合作用下的气井持续环空压力预测模型[J]. 中国安全科学学报, 2023, 33(2): 166-172.
[15]	朱彤, 秦丹, 魏雯, 任杰, 冯移冬. 基于机器学习的公交驾驶员事故风险识别及影响因素研究[J]. 中国安全科学学报, 2023, 33(2): 23-30.