深井动压巷道围岩移近量的BP神经网络预测

Displacement Forecast of Rockstrata around Deep Roadway under Dynamic Pressure Utilizing Back-propagation Neural Network

传统的预测深井动压巷道围岩移近量的方法主要有:解析法、数值分析法、回归分析法、概率分析法、模糊分析法等。文章主要介绍BP神经网络对深井动压巷道围岩移近量的预测、原理及其Matlab的程序实现。通过分析表明,BP神经网络在深井动压巷道围岩移近量的预测中具有较高的精度。

Traditional methods of forecasting displacement of rockstrata around deep roadway under dynamic pressure are mainly： analytical method, numerical analytical method, regressive analytical method, probabilistic analytical method, fuzzy analytical method, etc. The paper mainly introduces how to apply Back - propagation neural network to forecasting displacement of rockstrata around deep roadway under dynamic pressure, the principle of Back- propagation neural network and its Matlab Procedure algorithm. Through analysis, the result proves to be of upper precision applying Back - propagation neural network to forecasting displacement of rockstrata around deer） roadwav under dvnamic pressure.