摘要
如何科学、合理地预测路基沉降是高速公路建设的关键,针对传统的BP神经网络算法易陷入局部极小值,影响预测精度的问题,本研究利用遗传算法(genetic algorithm,GA)的全局寻优能力对BP神经网络进行优化,构建了基于GA-BP神经网络的路基沉降预测模型.以高速公路路基沉降监测进行实验验证分析,结果表明:相较于传统BP神经网络模型,GA-BP神经网络模型的预测精度有了显著的提高,可为高速公路的建设及后续工程的沉降预测提供参考.
Scientific and reasonable prediction of subgrade settlement is the key to expressway construction,and the traditional BP neural network algorithm is easy to fall into local minimum,which affects the prediction accuracy.In this study,Genetic algorithm(GA),having the global optimization ability,is used to optimize BP neural network,and a prediction model of subgrade settlement based on GA-BP neural network is constructed.The experimental verification analysis is carried out by monitoring the settlement of expressway subgrade.The experimental results show that compared with the traditional BP neural network model,the prediction accuracy of GA-BP neural network model is significantly improved,which can provide reference for expressway construction and follow-up projects.
作者
周卫
王倩
Zhou Wei;Wang Qian(The First Surveying and Mapping Institute of Hunan Province,Changsha Hunan 410114;Changsha University of Science&Technology,Changsha Hunan 410114)
出处
《国土资源导刊》
2022年第3期76-80,共5页
Land & Resources Herald
关键词
遗传算法
BP神经网络
路基沉降
沉降预测
genetic algorithm
BP neural network
subgrade settlement
settlement prediction
作者简介
第一作者:周卫(1989—),男,助理工程师,研究方向:工程测量与测绘基准建设,E-mail:352908131@qq.com。
