基于巷道摩擦阻力系数BP神经网络预测模型的矿井风网风量预测研究  

Study on the air quantity of mine ventilation network based on BP neural network prediction model of friction resistance coefficient in roadway

作　　者：刘彦青 LIU Yanqing(Mine Safety Technology Branch,China Coal Research Institute,Beijing 100013,China;State Key Laboratory of Coal Mining and Clean Utilization of Coal Resources,Beijing 100013,China)

机构地区：[1]煤炭科学技术研究院有限公司安全分院,北京100013 [2]煤炭资源高效开采和洁净利用国家重点实验室,北京100013

出　　处：《矿业安全与环保》2021年第2期101-106,共6页Mining Safety & Environmental Protection

基　　金：中国煤炭科工集团科技创新基金项目(2018-TD-ZD002);煤炭科学技术研究院有限公司科技发展基金项目(2020CX-II-22)。

摘　　要：针对矿井风网解算中待掘巷道摩擦阻力系数难以准确实测赋值的问题,构建了巷道摩擦阻力系数BP神经网络预测模型,以双柳煤矿各类巷道摩擦阻力系数实测数据作为训练样本进行学习训练,使预测模型的期望误差达到0.0001以下。利用该模型对尚未贯通的23(4)13回采工作面的巷道摩擦阻力系数进行预测,将预测结果代入基于斯考德—恒斯雷法风网解算方法构建的双柳煤矿通风网络解算模型中,分别对23(4)13回采工作面贯通后备用阶段和回采阶段全风网风量分布进行了解算,解算结果与现场实测结果之间相对误差小于8%。研究结果表明,利用基于BP神经网络算法的巷道摩擦阻力系数预测模型对待掘巷道摩擦阻力系数进行预测赋值,能够实现风网风量的准确解算。In view of the problem that it was difficult to accurately assign the friction resistance coefficient of the roadway to be excavated by measurement in the ventilation network calculation,the BP neural network prediction model of roadway friction resistance coefficient was constructed.Taking the measured data of friction coefficient of various roadways in Shuangliu Coal Mine as training samples for learning and training,the expected error of the prediction model was less than 0.0001.The model was used to predict the friction resistance coefficient of the mining roadway in 23(4)13 working face,which had not yet been completed.The predicted results were substituted into the ventilation network calculation model of Shuangliu Coal Mine,and this model was based on the Scott-Hinsley method of ventilation network calculation,the air quantity distribution of whole ventilation network in 23(4)13 working face had been completed during the mining period and the back-up period was calculated.The relative error between the calculation results and the field measurement results was less than 8%.The research results show that the BP neural network prediction model of friction resistance coefficient in roadway can be used to evaluate the friction resistance coefficient of the roadway to be excavated,which can realize the accurate calculation of ventilation network.

关 键 词：待掘巷道 巷道摩擦阻力系数 影响因素 BP神经网络 风网解算 风量预测 

分 类 号：TD725[矿业工程—矿井通风与安全] TD726