含裂隙预崩解炭质泥岩的强度特性预测  

Prediction of Strength Characteristics of Pre-disintegrating Carbonaceous Mudstone with Fissures

作　　者：杨禹 章赛泽 YANG Yu;ZHANG Saize(School of Civil and Environmental Engineering,Changsha University of Science&Technology,Changsha 410114,China)

机构地区：[1]长沙理工大学土木与环境工程学院,湖南长沙410114

出　　处：《西北工程技术学报(中英文)》2025年第1期45-52,共8页Ningxia Engineering Technology

基　　金：湖南省教育厅科学研究项目(22B0342);长沙理工大学实践创新与创业项目(CLSJCX23025)。

摘　　要：准确预测含裂隙炭质泥岩路堤的抗剪强度对确保其长期稳定性至关重要。分析了炭质泥岩裂隙的最大长度、最大宽度、数量,裂隙率,水质量分数,干密度,垂直压力等7个影响因素,运用粒子群算法与BP神经网络相结合,建立了PSO-BP神经网络的抗剪强度预测模型。结果表明,相较于传统的BP神经网络,PSO-BP模型的平均绝对百分比误差、平均绝对误差和均方根误差分别降低了66.17%,63.79%和58.78%。此外,模型预测结果表明,裂隙数量增加时,抗剪强度先增加后减小,最终趋于平稳;裂隙最大长度增加,使得抗剪强度持续上升并最终稳定;而裂隙最大宽度和裂隙率增加时,抗剪强度逐渐减小并趋于平稳。改进的PSO-BP预测模型优于BP神经网络,能有效提高预测精度,为炭质泥岩路堤的加固和维护措施提供重要依据,保障其长期稳定性。Accurate prediction of strength of fissure-containing carbonaceous mudstone embankments is crucial for ensuring long-term stability.Seven influencing factors,including the maximum length of fissures,maximum width of fissures,number of fissures,fissure rate,water mass fraction,dry density,and vertical pressure were analyzed.The particles swarm algorithm was combined with the BP neural network,resulting in PSO-BP-based prediction models for shear strength.The results show that the average absolute error,root mean square error,and average absolute percentage error of the PSO-BP shear strength prediction model decreased by 66.17%,63.79%and 58.78%,respectively,compared with that of the traditional BP neural network.In addition,the model prediction results indicate that when the number of fissures increases,the shear strength initially increases then decreases,and finally stabilizes;the increase of the maximum length of the fissures leads to a continuous increase in shear strength which tends to stabilize;whereas the increase of the maximum width of the fissures and fissure rate results in a gradual decrease in the shear strength,which also tends to stabilize.The improved performance of the PSO-BP prediction model over the BP neural network model significantly enhances prediction accuracy,providing a reliable foundation for the formulation of scientific reinforcement and maintenance strategies to ensure the long-term stability of carbonaceous mudstone embankment slopes.

关 键 词：炭质泥岩 裂隙 路堤 抗剪强度 PSO-BP神经网络 

分 类 号：U416[交通运输工程—道路与铁道工程]