无督神经网络在开采煤层自燃危险性预测中的应用

APPLICATION OF UNSUPERVISED NEURON NETWORK IN PREDICTING COAL SPON TANEOUS COMBUSTION

摘要: 通过对开采煤层自燃机理的分析,选出了主要的影响因素,采用了人工神经网络方法,通过对典型样本的学习,建立了开采煤层自燃危险性预测的无督神经网络模型。并用“未知”样本进行了验证,结果表明了该方法的准确性。

Abstract: By analysing the mechanism of coal spontaneous combustion,the several main factors are selected,and the neuron network prediction model of coal spontaneous combustion is built through learning this model with the samples.The validation of the model has been made on basis of the prediction results for the “unknown”samples.