摘要
为了更加科学、准确地对煤层的自燃危险性进行预测,从煤自然属性、地质条件和开采条件3方面选取指标构建了煤层自燃危险性预测指标体系,结合遗传投影寻踪算法和聚类方法建立了煤层自燃危险性预测的遗传投影寻踪聚类模型。该模型通过运用遗传算法对投影方向进行优化,计算能够反映煤层自燃危险程度的一维投影特征值,并对投影特征值进行聚类分析,从而得出煤层自燃危险性等级。将模型运用于所选的8个样本工作面,得出的煤层自燃危险性预测等级和实际情况吻合度较高。结果表明,遗传投影寻踪聚类方法能够实现对工作面煤层自燃危险性程度的有效预测。
In order to improve the accuracy of predicting the spontaneous combustion danger of coal seam,a prediction index system of coal spontaneous combustion danger is established based on the natural properties of coal,geological conditions and mining conditions and combined with genetic projection pursuit and clustering method,the prediction model of coal seam spontaneous combustion risk is established with genetic projection pursuit clustering.The model calculates the one-dimensional projection characteristic value that can reflect the degree of spontaneous combustion of coal seam and carries out cluster analysis to get the rank of spontaneous combustion risk.The model is applied in the faces of the selected 8 samples,finding out that the prediction level is in sound agreement with the actual degree of spontaneous combustion of coal seam.The results show that the application of genetic projection pursuit clustering method is effective for the prediction of coal spontaneous combustion risk.
作者
陈雄
吕贵春
CHEN Xiong;LYU Guichun(Chongqing Vocational Institute of Engineering,Chongqing 402260)
出处
《工业安全与环保》
2018年第7期25-28,共4页
Industrial Safety and Environmental Protection
基金
国家自然科学基金(51574280)
国家"十三五"科技重大专项资助项目(2016ZX05045-004)
关键词
煤层自燃
危险性预测
遗传算法
投影寻踪
聚类方法
spontaneous combustion of coal seam
risk prediction
genetic algorithm
projection pursuit
cluster method
