CAJ | 학술논문

基于影响边坡稳定性各因素之间具有一定的相关性和边坡工程是一个非线性、不确定的动态过程等这些特征，首次应用主成分和BP神经网络的原理和方法，建立了边坡稳定性评价模型，并应用SPSS软件对影响因素进行分析并确定主成分，应用Matlab7.1神经网络工具箱对一些边坡样本进行训练仿真。对比了经过主成分分析和未经过主成分分析评价结果，结果表明，经过主成分分析的BP神经网络评价精度更高，相对误差更小。表明了建立主成分和BP神经网络评价模型具有较好的可行性和适用性。
기우영향변파은정성각인소지간구유일정적상관성화변파공정시일개비선성、불학정적동태과정등저사특정，수차응용주성분화BP신경망락적원리화방법，건립료변파은정성평개모형，병응용SPSS연건대영향인소진행분석병학정주성분，응용Matlab7.1신경망락공구상대일사변파양본진행훈련방진。대비료경과주성분분석화미경과주성분분석평개결과，결과표명，경과주성분분석적BP신경망락평개정도경고，상대오차경소。표명료건립주성분화BP신경망락평개모형구유교호적가행성화괄용성。
Based on the characteristics of a certain correlation between various factors affecting the slope stability and the slope engineering is a dynamic process being nonlinear and uncertain , the methods of principal component analysis and BP neural network were applied to set up the evaluation model of slope stability .The influencing fac-tors were analyzed to determine the principal component by using SPSS software , and the network was then trained and simulated based on slope samples data set by using the NN toolbox in Matlab 7.1.The results with principal component analysis and without principal component analysis were compared .It showed that the former is better , with higher accuracy and smaller relative error , the method of principal component analysis and BP neural network established in this paper has a good feasibility and applicability .