基于机器学习的边坡稳定性预测方法研究  

Research on slope stability prediction method based on machine learning

作　　者：肖克华[1] XIAO Kehua(Hubei Coal Geological and Geophysical Survey Team,Wuhan 430200,Hubei,China)

机构地区：[1]湖北煤炭地质物探测量队,湖北武汉430200

出　　处：《矿产勘查》2024年第S2期85-92,共8页Mineral Exploration

摘　　要：为了快速、可靠、准确地评价边坡稳定性,提出了一种基于优化算法优化BP神经网络的边坡稳定性预测方法,该方法可以通过基本的几何、地质因素客观地评价边坡稳定性,克服了传统机器学习模型选择困难和误判风险高的缺点。收集505组边坡样本,将优化算法优化的BP神经网络预测模型与常用的学习模型进行预测性能比较。研究结果表明:基于同样的数据集,思维进化算法(MEA)优化BP神经网络模型泛化性能最佳。MEA-BP神经网络模型能够有效地预测边坡的稳定性,精确率P为0.903,ROC曲线线下面积(AUC)为0.93,在预测的准确性及泛化能力上明显优于BP神经网络,基于MEA-BP的方法有望成为边坡稳定性预测的通用方法。In order to quickly,reliably,and accurately evaluate slope stability,a slope stability prediction method based on optimization algorithm optimized BP neural network is proposed.This method can objectively evaluate slope stability through basic geometric and geological factors,overcoming the shortcomings of traditional machine learning model selection difficulties and high risk of misjudgment.Collect 505 sets of slope samples and compare the prediction performance of the optimized BP neural network prediction model with commonly used learning models.The research results indicate that based on the same dataset,the Mind Evolutionary Algorithm(MEA)optimizes the generalization performance of the BP neural network model with the best performance.The MEA-BP neural network model can effectively predict the stability of slopes with an accuracy rate of P of 0.903 and an area below the ROC curve(AUC)of 0.93.It is significantly superior to the BP neural network in terms of prediction accuracy and generalization ability.The method based on MEA-BP is expected to become a common method for predicting slope stability.

关 键 词：边坡稳定性 思维进化算法 BP神经网络 

分 类 号：P642.22[天文地球—工程地质学]