摘要
对小型水库大坝及岸坡变形进行准确预测是水库现代化管理工作的重要环节。为提高小型水库大坝及岸坡变形预测的准确性,基于PS-InSAR技术与M-LSTM神经网络(多变量长短记忆法),提出小型水库大坝及岸坡变形预测方法,首先利用PS-InSAR技术获取浏阳市4座典型小型水库坝体及坡岸变形特征,然后优化出3种变形影响因素,建立基于M-LSTM的小型水库大坝及岸坡变形预测模型,并对其准确性进行了验证。结果表明,PS-InSAR技术在小型水库坝体及坡岸变形监测中具有良好的可操作性;M-LSTM模型比LSTM模型具有更好的预测效果,其平均判定系数达到了0.91,平均绝对误差、平均均方根误差也仅为0.012、0.010,可见M-LSTM模型在小型水库大坝及岸坡变形预测中有较好的适用性。
Accurate prediction of dam and bank slope deformation of small reservoirs is an important link of reservoir modernization management.This paper proposes a method for predicting the deformation of small reservoir dams and slopes based on PS-InSAR technology and M-LSTM neural network(Multivariate Long Short-Term Memory).Firstly,the PS-InSAR technology was used to obtain the deformation characteristics of four typical small reservoir dams and slopes in Liuyang City.Then,three deformation influencing factors were optimized to establish a deformation prediction model of small reservoir dams and slopes based on M-LSTM.The accuracy of the model was verified.The results show that the PS-InSAR technology has good operability in monitoring the deformation of small reservoir dams and slopes.The M-LSTM model has better prediction performance compared to the LSTM model,with an average coefficient of determination reaching 0.91.The average absolute error and root mean square error are only 0.012 and 0.010,respectively,indicating the good applicability of the M-LSTM model in predicting the deformation of small reservoir dams and slopes.
作者
王祥
姜楚
蒋煌斌
WANG Xiang;JIANG Chu;JIANG Huang-bin(Hunan Institute of Water Resources and Hydropower Research,Changsha 410007,China;Hunan Dam Safety and Disease Prevention Engineering and Technology Research Centre,Changsha 410007,China)
出处
《水电能源科学》
北大核心
2024年第10期153-157,共5页
Water Resources and Power
基金
湖南省自然科学青年基金项目(2024JJ6282)
湖南省大坝安全与病害防治工程研究中心项目(Hndam2023kf05)。
