改进型小波神经网络在高铁路基沉降精准监测中的应用研究  

作　　者：刘攀 王倩 

机构地区：[1]甘肃建筑职业技术学院,兰州730050

出　　处：《科技创新与应用》2025年第9期64-68,共5页Technology Innovation and Application

基　　金：2023年高校教师创新基金项目(2023A-264)。

摘　　要：鉴于传统BP神经网络在处理复杂非线性问题时存在的局限性,该研究致力于对广泛应用的小波神经网络模型实施优化策略。具体而言,通过将高速铁路路基沉降数据以时间序列的形式作为输入向量,构建一个增强的预测框架,旨在精确预估路基的沉降量。实验验证环节显示,相较于未改进的小波神经网络模型,该优化后的小波神经网络在沉降量预测任务中展现出更优的性能,提升预测结果的精度与可靠性,从而为高速铁路基础设施的安全监测与维护提供技术支持。In view of the limitations of traditional BP neural networks in dealing with complex nonlinear problems,this research is committed to implementing optimization strategies for widely used wavelet neural network models.Specifically,by using high-speed railway subgrade settlement data in the form of time series as input vectors,an enhanced prediction framework is constructed to accurately predict subgrade settlement.The experimental verification process shows that compared with the unimproved wavelet neural network model,the optimized wavelet neural network shows better performance in the settlement prediction task,improving the accuracy and reliability of the prediction results,thereby providing technical support for the safety monitoring and maintenance of high-speed railway infrastructure.

关 键 词：路基沉降 改进小波神经网络 沉降量预测 实验验证 预测框架 

分 类 号：U213.157[交通运输工程—道路与铁道工程]