论文部分内容阅读
煤层自燃温度与气体产物之间存在对应关系,据此可预报煤层自燃。但是,温度与气体产物之间的关系为非线性的,非常复杂,不能用简单的数学公式表达。用前向3层人工神经网络模型表征这一对应关系,并用大量实验数据对网络进行训练以获得神经元间的连接强度。据此建立了人工神经网络专家系统,可以根据现场监测的煤自燃指标气体等参数,判断煤自燃现状,并预测自燃发展趋势。实践表明,该神经网络专家系统能够较准确地预报煤层自燃。
Coal spontaneous combustion temperature and gas products exist between the corresponding relationship, which can predict coal spontaneous combustion. However, the relationship between temperature and gas products is non-linear, very complex and can not be expressed using simple mathematical formulas. This three-layer artificial neural network model is used to characterize this correspondence and a large number of experimental data are used to train the network to obtain the connection strength between neurons. Based on this, an artificial neural network expert system is established, which can judge the status of spontaneous combustion of coal and predict the development trend of spontaneous combustion according to the parameter of spontaneous combustion index gas monitored on site. Practice shows that the neural network expert system can predict coal spontaneous combustion more accurately.