硫化矿石自燃灾害预警的RBF神经网络模型及应用  

Early Warning of Spontaneous Combustion of Sulfide Ore and Its Application Based on RBF Neural Network

作　　者：蔡逸伦 阳富强[1] 刘晓霞[1] CAI Yilun;YANG Fuqiang;LIU Xiaoxia(College of Environment and Resources,Fuzhou University,Fuzhou 350116,China)

机构地区：[1]福州大学环境与资源学院

出　　处：《有色金属工程》2019年第7期72-78,共7页Nonferrous Metals Engineering

基　　金：国家自然科学基金资助项目(51874100,51741402);河南省瓦斯地质与瓦斯治理重点实验室-省部共建国家重点实验室培育基地开放基金资助项目(WS2017B03)~~

摘　　要：为了科学准确地确定硫化矿石自燃灾害预警等级,进而为灾害防控提供决策,减少矿山损失。通过深入分析硫化矿石自燃典型案例,按照预警指标选取原则,从矿山生产人员、硫化矿石自燃倾向性、环境条件、管理水平等4个方面构建硫化矿石自燃灾害预警系统指标体系。运用RBF预警模型对硫化矿石自燃灾害预警等级进行预测,采用所选的21组样本数据完成了RBF神经网络的学习与训练。应用学习好的预警模型对江西某高硫矿山的硫化矿石自燃灾害预警等级进行预测。该矿山硫化矿石自燃灾害预警等级为Ⅲ级,即自燃危险性一般,与该采场的实际状况相一致。通过现场案例验证了该模型的适用性,能应用于硫化矿石自燃灾害预警等级的预测,对类似灾害事件预警也有借鉴作用。In order to scientifically and accurately determine the early warning level of spontaneous combustion hazard of sulfide ore, reduce the loss of mine disaster and provide decision-making for disaster early warning. Through in-depth analysis of the typical cases of spontaneous combustion of sulfide ore, according to the early warning index selection, the index system of early warning system for spontaneous combustion disaster of sulfide ore was constructed from four aspects of mine production personnel, spontaneous combustion tendency of sulfide ore, environmental conditions, and management level. The RBF early warning model was used to predict the warning level of spontaneous combustion hazard of sulfide ore, and 21 sets of sample data selected were used to complete the learning and training of RBF neural network. Then the well-preserved early warning model was used in a high-sulfur mine in Jiangxi. The result showed that the warning level of spontaneous combustion disaster of sulfide ore in this mine is grade III and the risk of spontaneous combustion is generally the same, which is consistent with the actual situation of the stope. The field case shows the applicability of the model. Therefore the model can be applied to the prediction of early warning grade of spontaneous combustion disaster of sulfide ore, and it can also be used as a reference for early warning of similar disaster events.

关 键 词：硫化矿石 自燃 RBF神经网络 预警模型 

分 类 号：TD752[矿业工程—矿井通风与安全]