Prediction model of coal spontaneous combustion risk based on PSO-BPNN

doi:10.16265/j.cnki.issn1003-3033.2023.07.2235

Abstract

Abstract:

In order to improve the accuracy of the risk prediction of coal spontaneous combustion in the goaf, a PSO algorithm was used to improve the connection weight and threshold values of BPNN. A coal spontaneous combustion risk prediction model (PSO-BPNN model) coupling the PSO algorithm with BPNN was constructed. The PSO-BPNN model was compared and analyzed with the prediction results of the standard BPNN model and SVR model respectively. The results show that the optimized mean relative error, mean absolute error and root mean square error are reduced by 9.35%, 0.170 7 and 0.205 6, respectively, and the coefficient of determination is increased by 0.116 9 compared with the BPNN model. Compared with the SVR model, the optimized mean relative error, mean absolute error and root mean square error are reduced by 5.41%, 0.115 2 and 0.171 5, respectively, and the coefficient of determination is increased by 0.0891. It was demonstrated that the PSO-BPNN model has higher prediction accuracy than the standard BPNN model and SVR model.

Key words: coal spontaneous combustion, particle swarm optimization(PSO), back propagation neural network(BPNN), support vector regression(SVR), prediction model, goaf

WANG Wei, LIANG Ran, QI Yun, JIA Baoshan, WU Zewei. Prediction model of coal spontaneous combustion risk based on PSO-BPNN[J]. China Safety Science Journal, 2023, 33(7): 127-132.

Figures/Tables 11

Fig.1

Tab.1

Fig.2

Fig.3

Tab.2

Fig.4

Fig.5

Tab.3

Tab.4

Tab.5

Fig.6

References 12

[1]	QIAO Ming, REN Ting, ROBERTS J, et al. New insight into proactive goaf inertisation for spontaneous combustion management and control[J]. Process Safety and Environmental Protection, 2022, 161(4):739-757. doi: 10.1016/j.psep.2022.03.074
[2]	周亮, 戴广龙, 秦汝祥. 高瓦斯易自燃煤层采空区遗煤自燃影响因素研究[J]. 中国安全科学学报, 2018, 28(2): 122-127. doi: 10.16265/j.cnki.issn1003-3033.2018.02.021
	ZHOU Liang, DAI Guanglong, QIN Ruxiang. Research on factors influencing spontaneous combustion of coal in goaf in gassy coal seam having propensity to spontaneous combustion[J]. China Safety Science Journal, 2018, 28 (2):122-127. doi: 10.16265/j.cnki.issn1003-3033.2018.02.021
[3]	常绪华, 罗乾坤. 基于改进博弈论采空区煤自燃危险性可拓评价[J]. 矿业安全与环保, 2022, 49(4):211-218.
	CHANG Xuhua, LUO Qiankun. Coal spontaneous combustion hazard extension evaluation in goaf based on improved game theory[J]. Mining Safety & Environmental Protection, 2022, 49(4):211-218.
[4]	汪伟, 贾宝山, 祁云. 改进CRITIC修正G2-TOPSIS的钻孔自燃预测模型及应用[J]. 中国安全科学学报, 2019, 29(11):26-31. doi: 10.16265/j.cnki.issn1003-3033.2019.11.005
	WANG Wei, JIA Baoshan, QI Yun. Prediction model of spontaneous combustion risk of extraction drilling based on improved CRITIC modified G2-TOPSIS method and its application[J]. China Safety Science Journal, 2019, 29 (11):26-31. doi: 10.16265/j.cnki.issn1003-3033.2019.11.005
[5]	祁云, 齐庆杰, 汪伟, 等. 基于集值统计-Entropy的采空区自燃危险性预测模型及应用[J]. 安全与环境学报, 2019, 19(5):1526-1531.
	QI Yun, QI Qingjie, WANG Wei, et al. A forecast model for the spontaneous combustion risk in the goaf based on set valued statistics-entropy and its application[J]. Journal of Safety and Environment, 2019, 19(5):1526-1531.
[6]	孟倩. 基于计算智能的煤矿自然发火预警研究[D]. 徐州: 中国矿业大学, 2015.
	MENG Qian. Study on pre-warning of spontaneous fire in coal mine based on computational intelligence[D]. Xuzhou: China University of Mining and Technology, 2015.
[7]	邢媛媛, 张飞飞. 基于AEM-TOPSIS的煤炭自燃风险评价研究[J]. 煤炭工程, 2021, 53(11):131-134.
	XING Yuanyuan, ZHANG Feifei. Risk evaluation of coal spontaneous combustion based on AEM-TOPSIS[J]. Coal Engineering, 2021, 53(11):131-134.
[8]	邓念东, 丁一, 邢聪聪, 等. 基于PCA-BP神经网络的风积砂充填体强度预测[J]. 矿业研究与开发, 2021, 41 (2):109-113.
	DENG Niandong, DING Yi, XING Congcong, et al. Strength prediction of eolian sand filling body based on PCA-BP neural network[J]. Mining Research & Development, 2021, 41(2):109-113.
[9]	林海飞, 高帆, 严敏, 等. 煤层瓦斯含量PSO-BP神经网络预测模型及其应用[J]. 中国安全科学学报, 2020, 30(9):80-87.
	LIN Haifei, GAO Fan, YAN Min, et al. Study on PSO-BP neural network prediction method of coal seam gas content and its application[J]. China Safety Science Journal, 2020, 30(9):80-87.
[10]	朱红青, 常文杰, 张彬. 回采工作面瓦斯涌出BP神经网络分源预测模型及应用[J]. 煤炭学报, 2007, 32 (5): 504-508.
	ZHU Hongqing, CHANG Wenjie, ZHANG Bin. Different-source gas emission prediction model of working face based on BP artificial neural network and its application[J]. Journal of China Coal Society, 2007, 32 (5):504-508.
[11]	梁盛开, 曹琼, 罗杨阳. 煤矿瓦斯突出的神经网络预测[J]. 固体力学学报, 2010, 31(增1):180-183.
	LIANG Shengkai, CAO Qiong, LUO Yangyang. Prediction of coal mine's coal and gas outburst based on neural network[J]. Chinese Journal of Solid Mechanics, 2010, 31(S1):180-183.
[12]	鲁学富. 基于GA-SVM的采空区煤自燃预警技术研究及应用[D]. 西安: 西安科技大学, 2020.
	LU Xuefu. Research and application on early waring technology of coal spontaneous combustion in goaf based on GA-SVM[D]. Xi'an: Xi'an University of Science and Technology, 2020.

序号	ϕ(O₂)/%	ϕ(N₂)/%	…	ϕ(CO₂/CO)	危险等级
1	20.35	79.58	…	7.14	1
2	20.4	79.47	…	3.12	1
3	20.19	79.65	…	1.75	1
4	20.57	79.35	…	2.6	1
5	20.31	79.51	…	3.94	1
︙	︙	︙		︙	︙
46	20.62	78.81	…	2.33	1
47	19.89	79.05	…	7.48	1
48	11.33	79.76	…	0.8	3
49	11.9	78.86	…	0.8	2
50	16.71	81.09	…	4.03	1

序号	真实值	预测值	绝对误差	相对误差/%
1	3	2.815 4	0.184 6	6.15
2	3	2.943 1	0.056 9	1.90
3	2	1.909 5	0.090 5	4.52
4	1	0.815 5	0.184 5	18.45
5	2	2.619 5	0.619 5	30.98
6	1	1.024 2	0.024 2	2.42
7	1	1.202 7	0.202 7	20.27
8	3	2.546 1	0.453 9	15.13
9	2	1.612 7	0.387 3	19.36
10	1	1.181 1	0.181 1	18.11

序号	真实值	预测值	绝对误差	相对误差/%
1	3	2.989 5	0.010 5	0.35
2	3	3.016 4	0.016 4	0.55
3	2	1.940 5	0.059 5	2.98
4	1	0.863 5	0.136 5	13.65
5	2	1.795 8	0.204 2	10.21
6	1	0.988 5	0.011 5	1.15
7	1	0.986 5	0.013 5	1.35
8	3	2.965 8	0.034 2	1.14
9	2	2.134 9	0.134 9	6.75
10	1	0.943 1	0.056 9	5.69

序号	真实值	预测值	绝对误差	相对误差/%
1	3	2.494 1	0.505 9	16.86
2	3	3.083 8	0.083 8	2.79
3	2	1.444 8	0.555 2	27.76
4	1	1.109 3	0.109 3	10.93
5	2	2.044 9	0.044 9	2.25
6	1	1.025 4	0.025 4	2.54
7	1	0.998 2	0.001 8	0.18
8	3	2.781 6	0.218 4	7.28
9	2	1.975 2	0.024 8	1.24
10	1	1.260 3	0.260 3	26.03

模型	性能指标
模型	平均绝对误差	平均相对误差/%	均方根误差	判定系数
BPNN	0.238 5	13.73	0.299 0	0.870 5
PSO-BPNN	0.067 8	4.38	0.093 4	0.987 4
SVR	0.183 0	9.79	0.264 9	0.898 3