硫化矿石自燃倾向性等级划分的GA-BP神经网络模型及应用

发布时间：2018-04-27 13:05

本文选题：硫化矿石 + 自燃倾向性　；参考：《自然灾害学报》2015年04期

【摘要】：准确判定硫化矿石的自燃倾向性等级,可为高硫矿井防灭火设计和安全回采提供重要的理论依据。结合已有的研究成果,选取表征硫化矿石自燃倾向性本质特性的自热点温度、室内低温氧化质量增加率和自燃点温度3项指标作为硫化矿自燃倾向性等级划分的基本判别因子。以采自典型高硫矿山的18组矿样的实测数据作为训练样本进行分析计算,进而建立了硫化矿石自燃倾向性等级划分的GA-BP神经网络模型。最后,运用该模型对6组硫化矿井矿样的自燃倾向性进行了分级预测。研究结果表明,6组实测样本的误判率平均值仅为2.5%,取得了良好的预测效果。因此,GABP神经网络模型可以用于指导高硫矿山矿石自燃倾向性等级的划分。
[Abstract]:Accurate determination of spontaneous combustion tendency of sulphide ore can provide important theoretical basis for fire prevention design and safe mining of high sulfur mine. Combined with the existing research results, three indexes, namely, the autoignition temperature, the increase rate of indoor low temperature oxidation mass and the spontaneous combustion point temperature, are selected as the basic discriminant factors for the classification of the spontaneous combustion tendency of sulfide ores. The GA-BP neural network model for the classification of spontaneous combustion tendency of sulphide ores was established by using the measured data of 18 groups of ore samples collected from typical high-sulfur mines as training samples for analysis and calculation. Finally, the model is used to predict the spontaneous combustion tendency of 6 groups of sulphide mine samples. The results show that the average misjudgment rate of 6 groups of samples is only 2.5, and good prediction results have been obtained. Therefore, the GABP neural network model can be used to guide the classification of spontaneous combustion tendency of ore in high sulfur mines.
【作者单位】：福州大学环境与资源学院;
【基金】：国家自然科学基金项目(51304051) 福建省自然科学基金项目(2012J05088) 福州大学科技发展基金项目(2013-XQ-18) 福州大学科研启动基金项目(XRC-1174)
【分类号】：TD752.2

【参考文献】