改进BP算法在煤与瓦斯突出预测中的应用

为了正确预测煤与瓦斯突出的趋势与危险性 ,基于反向BP神经网络 ,笔者提出了一种改进的BP网络模型 :为了加快BP网络的收敛速度 ,增强其跳出局部极小点的能力 ,采用了自适应变步长法和改进模拟退火法 (SA法 )相结合的方法。实际应用表明 ,该模型收敛速度快 ,准确性高 ,具有较高的可靠性和实用性 ,是一种十分有效的煤与瓦斯突出危险性预测方法。

Application of Modified BP Neural Network in Predicting Coal and Gas Outburst

In order to accurately predict the potentiality and risk of coal and gas outburst, a kind of modified BP neural network was suggested, in which the network convergence speed is quickened and falling into local minimum point is avoided, a combined method with changing network iteration step length and the simulated annealing (SA) is adopted. Practical application demonstrates that the prediction model has fast convergence speed and good prediction accuracy, which is very efficient in predicting the coal and gas outburst, and meaningful in the mine production safety.