摘要
为解决混凝土重力坝变形预测中传统监控模型存在的预测精度受限问题,提出了一种基于斑马优化算法(ZOA)和双向长短记忆网络(BiLSTM)的大坝变形预测模型。首先,挑选变形区域内典型的测点数据;随后,通过ZOA优化算法对模型超参数进行迭代优化,以提高模型的性能;最终,结合提出的ZOA-BiLSTM预测流程,实现了对混凝土重力坝变形的高精度预测。工程实例表明,模型预测结果与坝体变形空间分布特征相吻合。为监控大坝整体安全性态提供了一种新的、有效的方法。
To address the limitations in prediction accuracy inherent in traditional monitoring models for deformation prediction of concrete gravity dams,this study proposes a dam deformation prediction model based on the Zebra Optimization Algorithm(ZOA)and Bidirectional Long Short-Term Memory(BiLSTM)network.Initially,typical measurement point data within the deformation zone are selected.Subsequently,the hyperparameters of the model are iteratively optimized using the ZOA to enhance model performance.Finally,a high-precision prediction of concrete gravity dam deformation is achieved through the proposed ZOA-BiLSTM prediction workflow.Engineering case study demonstrates that the model s predictions align with the spatial distribution characteristics of dam deformation,offering a novel and effective methodology for monitoring the comprehensive safety status of dams.
作者
黄建生
周站勇
林兴铖
黄之源
林舒婷
赵晨博
HUANG Jiansheng;ZHOU Zhanyong;LIN Xingcheng;HUANG Zhiyuan;LIN Shuting;ZHAO Chenbo(Huadian Fuxin Zhouning Pumped Storage Co.,Ltd.,Ningde 355400,Fujian,China;College of Water Conservancy&Hydropower Engineering,Hohai University,Nanjing 210024,Jiangsu,China)
出处
《水力发电》
2025年第5期112-118,124,共8页
Water Power
基金
国家自然科学基金资助项目(52309151)
国家重点研发计划(2022YFC3005403)。
