露天矿最终边坡角的BP神经网络预测及数值模拟验证

利用专家调查表的方式分析露天矿山边坡稳定性的主要影响因素,确定网络模型的输入层神经元。研究BP神经网络的结构和输出方式,对神经网络模型进行优化,并收集大量矿山稳定边坡的相关参数作为样本库,建立神经网络预测模型,将该模型用于某大型深凹露天矿山最终边坡角的预测。根据预测边坡角参数,采用SURPAC—MIDAS—FLAC3D多软件组合的方式,创建逼真的矿山边坡稳定性三维数值计算模型并进行模拟分析,结果表明,模型预测的边坡角满足设计要求。

BP Neural Network Prediction and Verification with Numerical Simulation in Ultimate Pit Slope of Strip Mine

By the method of expert questionnaire,main factors which impact slope stability of open pit mine are analyzed,and the input layers of the network model are determined.BP neural network structure and output mode are studied,and neural network model has been optimized.With the sample database contains a large number of parameters collected from existing mines with stable slope,a neural network prediction model is established which is used to predict the final slope angle of a open pit mine.According to the slope angle parameters,using multi-software combination including SURPAC,MIDAS and FLAC3D to create a three-dimensional numerical calculation mode of mine slope stability and simulation analysis,the result shows that the slope angle predicted by the BP neural network model meets the design requirement.