摘要
综合利用主成分分析法(Principle Component Analysis,PCA)对影响公路岩质边坡稳定性的6个因素进行主成分提取,提取的 4 个主成分作为 BP 神经网络的输入变量,边坡状态作为输出变量,并采用Levenberg-Marquardt(LM)算法优化BP神经网络,建立了基于PCA-LM-BP 神经网络的公路边坡稳定预测模型. 结合中南公路岩质边坡工程实例,将PCA-LM-BP神经网络模型的预测结果与LM-BP神经网络模型、BP神经网络模型预测结果进行对比. 结果表明:基于PCA-LM-BP神经网络的预测模型精度较高,可为预测中南公路边坡稳定性提供一定的依据.
We extracted six factors that affect the stability of highway rock slope using the principal component analysis method(PCA).On the basis of taking the extracted four principal components as the input variables,and the state of slope as the output variables of BP neural network,we established a prediction model of highway slope stability based on PCA-LM-BP neural network by use of the Levenberg Marquardt algorithm.Taking the rock slope engineering of Zhongnan highway as an example,we compared the prediction results of PCA-LM-BP neural network model with those of LM-BP and BP neural network models.The results show that the prediction model based on PCALM-BP neural network has a higher accuracy,enabling it to serve as a criterion for the prediction of the slope stability of Zhongnan Highway.
作者
牛鹏飞
周爱红
NIU Pengfei;ZHOU Aihong(School of Exploration Technology and Engineering,Hebei GEO University,Shijiazhuang 050031,China;Hebei Center for Ecological and Environmental Geology Research,Hebei GEO University,Shijiazhuang 050031,China)
出处
《防灾科技学院学报》
2020年第1期10-16,共7页
Journal of Institute of Disaster Prevention
基金
国家自然科学基金资助项目(41807231)
河北省教育厅在读研究生创新能力培养资助项目(CXZZSS2019115)
河北地质大学第十六届学生科技基金后补助科研项目(KAD201906)。
作者简介
牛鹏飞(1995-),男,硕士研究生,研究领域为地质灾害防治;通讯作者:周爱红(1976-),女,博士,教授,研究领域为岩土工程、地质灾害.
