CAJ | 학술논문

基于BP神经网络分析方法,选取合理的边坡物理参数、力学参数和几何参数,建立露天矿边坡稳定性分析模型.按照不同的分析目的与要求,分别建立了BP神经网络可行性模型、滑坡面预测模型和治理滑坡面模型;选取分析样本,进行网络结构及学习参数的优化;选取最优的学习样本进行学习训练,据学习结果,利用勘测数据对目标进行预测,其预测结果能够满足精度要求.为了进一步证明其可行性引入模糊的方法,通过信息扩散化原理进行二次数值计算,所得计算结果与之前预测结果吻合.
기우BP신경망락분석방법,선취합리적변파물리삼수、역학삼수화궤하삼수,건립로천광변파은정성분석모형.안조불동적분석목적여요구,분별건립료BP신경망락가행성모형、활파면예측모형화치리활파면모형;선취분석양본,진행망락결구급학습삼수적우화;선취최우적학습양본진행학습훈련,거학습결과,이용감측수거대목표진행예측,기예측결과능구만족정도요구.위료진일보증명기가행성인입모호적방법,통과신식확산화원리진행이차수치계산,소득계산결과여지전예측결과문합.
Based on BP neural network, selecting reasonable physical parameters, mechanical parameters and geometrical parameters of the side slope, an open pit slope stability analysis model was established. According to different analytical purposes and requirements, BP neural network model, landslide surface model and landslide surface model are established; selecting the optimal learning samples for learning and training, according to study results, the target was predicted by using the survey data, and the prediction results can meet the accuracy requirements. To further prove its feasibility, the fuzzy method was introduced; according to the principle of information diffusion, the second numerical calculation was done, the results are consistent with the previously predicted results.