AGO-BP神经网络在主要影响角正切求取中的应用
Application of AGO-BP Neural Network to the Tangent Calculation Method of the Main Influencing Angle
-
摘要: 主要影响角正切tanβ是用概率积分法进行开采沉陷预计的主要参数之一,决定着地表沉陷的影响范围。指出了影响主要影响角正切的主要地质采矿因素,并根据一些矿区的实际观测资料,建立了求取主要影响角正切的AGO-BP神经网络模型。该模型是首先运用灰色理论中的累加算法对选定的原始计算数据进行预处理,然后采用BP神经网络模型计算主要影响角正切。AGO-BP神经网络模型不仅能够自动调整网络参数,而且避免只采用BP神经网络进行计算时可能出现的模型不稳定问题,所得到的主要影响角正切精度有一定的提高。Abstract: The tangent of the main influencing angle tanβ is one of the most important parameter for mining subsidence prediction with the probability integral method and it determines the range of surface subsidence.Based on analyzing the geologic and mining factors which affect the tangent of the main influencing angle,the paper constructed an AGO-BP neutral network model.This model preprocessed the selected original data on the basis of grey theory,and then calculated the tangent of the main influencing angle with BP neural network mode1.The AGO-BP neural network model not only can adjust the network parameters automatically,but also can avoid the instability problem compared with only using BP neural network model.The tangent of the main influencing angle prediction is more accuracy with the AGO-BP neural network model.