基于SPSS与ANN的汾西矿区煤巷围岩稳定性分类

Classification of Surrounding Rock Stability of Coal Roadways in Fenxi Coal Mining Area based on SPSS and ANN

汾西矿区煤巷现场条件复杂、支护困难,因此合理的巷道围岩稳定性分类对后期的支护设计具有重要意义。统计了多篇文献资料,利用SPSS软件建立围岩稳定性分类指标数据库,进行频数分析,以统计分析结果为依据,结合汾西矿区的实际情况确定出11个围岩稳定性分类指标。采用附加动量法改进BP神经网络,应用MATLAB软件建立ANN煤巷围岩稳定性识别模型,并选取46条汾西矿区煤巷作为样本对模型进行学习训练。将该模型应用于12条汾西矿区煤巷进行检验,识别准确率为91.7%。该模型具有较高的类型识别准确性,非线性映射效果较好,适用于汾西矿区煤巷围岩稳定性分类。

In Fenxi Coal Mining Area,the field conditions of coal roadways are complicated and they are difficult to support,therefore,the reasonable stability classification of roadway surrounding rock is of great significance for late support design. Based on a number of literature statistics,the database of the surrounding rock stability classification indexes was set up by SPSS software and frequency analysis was done. Based on the results of statistical analysis and according to the actual conditions in Fenxi Coal Mining Area,11 classification indexes of surrounding rock stability were determined. The BP neural network was improved with the additional momentum method,and the stability identification model of ANN coal roadway surrounding rock was established by MATLAB software,then the training and learning of the model was carried out by selecting46 coal roadways in this mining area as the samples. Finally,the model was tested in 12 coal roadways in this mining area,the identification accuracy of the results was 91. 7%. This indicated that this model had higher type identification stability and better nonlinear mapping effect,so it is suitable for the stability classification of coal roadways surrouding rock in Fenxi Mining Area.