Scholarlink

<p>Seepage significantly impacts the stability of earth and rockfill dams, making effective monitoring essential. This research introduces a novel PLS-BO-BiLSTM model that integrates Partial Least Squares (PLS) regression with Bidirectional Long Short-Term Memory (BiLSTM) networks and Bayesian Optimization (BO). The model is further optimized using Grey Wolf Optimization (GWO) to account for the lag effects of water depth and precipitation. The novelty of the model lies in its ability to effectively address multicollinearity while improving the prediction of nonlinear time-series data in complex seepage scenarios. Key results from multiple engineering case studies demonstrate the model’s high predictive accuracy (<InlineEquation ID="IEq1"> <InlineMediaObject> <ImageObject Color="BlackWhite" FileRef="24_2024_3625_Article_IEq1.gif" Format="GIF" Height="17" Rendition="HTML" Resolution="72" Type="Linedraw" Width="71" /> </InlineMediaObject> <EquationSource Format="TEX">\({\textrm{R}}^{2} &gt; 0.98\)</EquationSource> <EquationSource Format="MATHML"><math> <mrow> <msup> <mrow> <mtext>R</mtext> </mrow> <mn>2</mn> </msup> <mo>&gt;</mo> <mn>0.98</mn> </mrow> </math></EquationSource> </InlineEquation>), significantly reducing mean absolute errors and showing strong generalizability during sudden seepage events caused by heavy rainfall. These findings highlight the practical utility of the model for real-world dam safety monitoring.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Enhanced Earth and Rockfill Dam Seepage Forecasting via an Integrated PLS-BO-BiLSTM Approach: A Novel Model Incorporating Lag Effects and Optimization Algorithms

  • 期刊论文

  • 发表日期：

    2025年1月6日

查看全文
  • Zhiwen Xie,
  • Liang Chen

摘要

Seepage significantly impacts the stability of earth and rockfill dams, making effective monitoring essential. This research introduces a novel PLS-BO-BiLSTM model that integrates Partial Least Squares (PLS) regression with Bidirectional Long Short-Term Memory (BiLSTM) networks and Bayesian Optimization (BO). The model is further optimized using Grey Wolf Optimization (GWO) to account for the lag effects of water depth and precipitation. The novelty of the model lies in its ability to effectively address multicollinearity while improving the prediction of nonlinear time-series data in complex seepage scenarios. Key results from multiple engineering case studies demonstrate the model’s high predictive accuracy ( \({\textrm{R}}^{2} > 0.98\) R 2 > 0.98 ), significantly reducing mean absolute errors and showing strong generalizability during sudden seepage events caused by heavy rainfall. These findings highlight the practical utility of the model for real-world dam safety monitoring.