基于LSSA-BP神经网络的煤层瓦斯含量预测方法研究  

Research on coalbed gas content prediction method based on LSSA-BP neural network

作　　者：侯恩科[1,2] 荣统瑞 卫勇锋 夏冰冰 谢晓深 HOU Enke;RONG Tongrui;WEI Yongfeng;XIA Bingbing;XIE Xiaoshen(School of Geology and Environment,Xi’an University of Science and Technology,Xi’an 710054,China;Shaanxi Provincial Key Laboratory of Geological Safeguard for Green Coal Development,Xi’an 710054,China;Shaanxi Coal Huangling Mining Co.,Ltd.,Huangling 727307,China)

机构地区：[1]西安科技大学地质与环境学院,陕西西安710054 [2]陕西省煤炭绿色开发地质保障重点实验室,陕西西安710054 [3]陕西陕煤黄陵矿业有限公司,陕西黄陵727307

出　　处：《煤矿安全》2023年第11期55-61,共7页Safety in Coal Mines

基　　金：陕煤集团科研计划资助项目(2021SMHKJ-BK-J-01,2020SMHKJ-C-52)。

摘　　要：为提高煤层瓦斯含量预测的精准性和可靠性,提出基于Logistic混沌映射改进的麻雀搜索算法优化BP神经网络的煤层瓦斯含量预测模型(LSSA-BP模型)。先通过灰色关联分析法(GRA)筛选瓦斯含量的主控因素作为LSSA-BP预测模型的输入层节点数,后利用Logistic混沌映射初始化麻雀种群以增加种群多样性,再采用LSSA对BP神经网络的权值和阈值进行优化,解决了单一BP模型收敛速度慢和易陷入局部极小的问题;通过模型应用,将LSSA-BP、SSABP和BP模型的预测结果进行对比。结果表明:LSSA-BP预测模型的平均绝对误差(MAE)、均方误差(MSE)、均方根误差(RMSE)和平均绝对百分比误差(MAPE)分别为0.3469 m^(3)/t、0.1721 m^(3)/t、0.4149 m^(3)/t和27.4036%,均优于其他模型,提高了煤层瓦斯含量预测的准确性和稳定性。In order to improve the accuracy and reliability of coal seam gas content prediction,this paper proposed the coal seam gas content prediction model(LSSA-BP model)based on Logistic chaotic mapping improved sparrow search algorithm to optimize BP neural network.Firstly,the main control factors of gas content were selected using grey correlation analysis(GRA)as the node number of the input layer of the LSSA-BP prediction model.Then,the sparrow population was initialized by Logistic chaotic mapping to increase the diversity of the population.The problems of slow convergence rate and easy to fall into local minimum of single BP model are solved.Through model application,the prediction results of LSSA-BP,SSA-BP and BP models are compared.The results show that:the mean absolute error(MAE),mean square error(MSE),root mean square error(RMSE)and mean absolute percentage error(MAPE)of LSSA-BP prediction model were 0.3469 m^(3)/t,0.1721 m(3)/t,0.4149 m^(3)/t and 27.4036%,respectively,which were better than other models.The accuracy and stability of coal seam gas content prediction are improved.

关 键 词：煤层瓦斯含量 BP神经网络 麻雀搜索算法 LOGISTIC混沌映射 灰色关联分析 

分 类 号：TD712[矿业工程—矿井通风与安全]