摘要
道路路面是保障汽车正常行驶的基础,其结构的准确检测关乎着人们的生命安全。针对目前路面结构模量反演算法准确率低的问题,提出了一种基于落锤式弯沉仪和改进差分进化(DE)算法的路面结构模量反演方法。将改进差分进化算法与其他算法进行性能对比分析,结果表明,此算法的AUC值和F1值分别为0.986和0.973,均优于对比算法。随后,在模量反演对比分析中,所提反演方法对4层路面结构模量反演的平均均方根误差分别为6.32%、6.67%、6.45%、6.47%,显著优于对比方法。结果表明,所提路面结构模量反演方法具有有效性和实用性,有助于提高道路路面检测的准确性,为相关研究提供理论依据。
Road pavement serves as the foundation for ensuring the normal operation of vehicles,and its accurate structural detection is crucial for people's life safety.To address the issue of low accuracy in current pavement structure modulus backcalculation algorithms,a pavement structure modulus backcalculation method based on the falling weight deflectometer and an improved Differential Evolution(DE)algorithm has been proposed.A performance comparison analysis was conducted between the improved DE algorithm and other algorithms,revealing that the AUC value and F1 score of this algorithm reached 0.986 and 0.973,respectively,both outperforming the comparative algorithms.Subsequently,in the modulus backcalculation comparison analysis,the proposed backcalculation method demonstrated average root mean square errors of 6.32%,6.67%,6.45%,and 6.47%for modulus backcalculation of four-layer pavement structures,significantly outperforming the comparative methods.The results indicate that the proposed pavement structure modulus backcalculation method is effective and practical,contributing to enhancing the accuracy of road pavement detection and providing a theoretical basis for relevant research.
作者
高鹏
GAO Peng(Weifang Highway Development Group Co.,Ltd.,Weifang,Shandong 261057,China)
出处
《自动化应用》
2025年第16期137-139,142,共4页
Automation Application
关键词
落锤式弯沉仪
差分进化算法
路面结构
模量反演
falling weight deflectometer
DE algorithm
pavement structure
modulus backcalculation
