基于PSO-SVM的山区营运高速公路边坡防治费用预测  

Prediction of slope prevention cost for mountainous operating expressway based on PSO-SVM

作　　者：肖秋明[1] 刘昕娆 XIAO Qiuming;LIU Xinrao(School of Traffic and Transportation Engineering,Changsha University of Science&Technology,Changsha 410114,China)

机构地区：[1]长沙理工大学交通运输工程学院,湖南长沙410114

出　　处：《长沙理工大学学报（自然科学版）》2022年第2期120-128,共9页Journal of Changsha University of Science and Technology:Natural Science

基　　金：国家自然科学基金资助项目(51878077);安徽省交通运输科技进步计划项目(201839);安徽省交通控股集团有限公司科技项目(AHJK-养-2019-0001)。

摘　　要：【目的】山区营运高速公路边坡的防治费用。【方法】以安徽省山区营运高速公路为研究对象,在综合考虑边坡基本情况、防治方案和价格因素3个方面共14个特征指标的基础上,建立了利用粒子群优化(particle swarm optimization,PSO)算法优化支持向量机(support vector machine,SVM)的山区营运高速公路边坡防治费用预测模型。通过PSO算法对SVM的惩罚因子和核函数参数进行优化,根据工程实例采用相对误差、均方根误差和判定系数等对所建模型的预测性能进行验证和评估,并与其他模型进行比较。【结果】应用所建模型预测山区营运高速公路的边坡防治费用,平均相对误差降低了41.9%,判定系数达到了0.953。【结论】所建模型具有较高的准确性和适用性,可为边坡防治决策提供参考。[Purposes]The paper aims to predict the slope prevention cost for mountainous operating expressway.[Methods]Taken the mountainous operating expressway in Anhui Province as the research object,based on the comprehensive consideration of 14characteristic indexes in 3aspects of basic situation,prevention scheme and price factor for the slope,aprediction model of the slope prevention cost for mountainous operating expressway was established by using the particle swarm optimization(PSO)algorithm to optimize the support vector machine(SVM).The penalty factor and kernel function parameters of SVM were optimized by the PSO algorithm.According to the engineering examples,the prediction performance of the established model was verified and evaluated by relative error,root mean square error and determination coefficient etc.,and compared with that of the other models.[Findings]The average relative error decreases by 41.9%,and the determination coefficient reaches0.953when using the established model to predict the slope prevention cost for mountainous operating expressway.[Conclusions]The model proposed in this paper has higher accuracy and applicability,and can provide some reference for decision-making of slope prevention.

关 键 词：高速公路 边坡防治 费用预测 支持向量机 粒子群优化算法 

分 类 号：U4-9[交通运输工程—道路与铁道工程]