Page 46 - 2025年第56卷第7期
P. 46
(2)本文构建的 AM-SSA-BiGRU 模型,通过 AM 动态调整影响因素的关注程度,增强对关键信息
的筛选能力,并利用 SSA 确定 BiGRU 的最佳超参数组合,解决了参数选择困难、效率低的问题。与
LSTM、BiGRU、AM-LSTM、AM-BiGRU 模型相比,R 最大提高 17.76%,MAPE、RMSE、MAE 分别最
2
大降低 93.67%、91.29%、92.66%,说明本文模型可有效表征和预测渗流变化趋势。
(3)实例分析表明,基于本文滞后分量量化方法的渗流预测模型对所有测点 UP3、UP6、UP9 和
UP13 渗流预测精度均有提升。该方法能够充分挖掘时序数据中库水位和温度与渗流之间的滞后关系,
从而显著提高渗流预测精度,为混凝土坝渗流预测提供了新的方法。
参 考 文 献:
[ 1 ] 姚可夫,田始光,漆一宁,等 . 高海拔区特征环境驱动下混凝土坝服役性能研究进展[J] 水利学报,2023,
.
54(6):717-728.
.
[ 2 ] 赵二峰,顾冲时 . 碾压混凝土坝安全服役关键技术研究进展[J] 水利水电科技进展,2022,42(1):11-20.
[ 3 ] WANG X L, YU H L, LV P, et al. Seepage safety assessment of concrete gravity dam based on matter-element
.
extension model and FDA[J] Energies,2019,12(3):502.
[ 4 ] ZHANG H Y,SONG Z Y,PENG P,et al. Research on seepage field of concrete dam foundation based on artificial
neural network[J] Alexandria Engineering Journal,2021,60(1):1-14.
.
[ 5 ] CHEN X D, XU Y, GUO H D, et al. Comprehensive evaluation of dam seepage safety combining deep learning
.
with dempster-shafer evidence theory[J] Measurement,2024,226:114172.
[ 6 ] HUANG H, CHEN B. Dam seepage monitoring model based on dynamic effect weight of reservoir water level[J]
.
Energy Procedia,2012,16:159-165.
.
[ 7 ] 吴中如 . 水工建筑物安全监控理论及其应用[M] 北京:高等教育出版社,2003.
[ 8 ] 刘应龙,傅蜀燕,盛韬桢,等 . 考虑滞后效应的复合土工膜土石坝渗流安全性态分析[J] 水电能源科学,
.
2015,33(10):46-49.
[ 9 ] LI M C, SI W, REN Q B, et al. An integrated method for evaluating and predicting long-term operation safety of
concrete dams considering lag effect[J] Engineering with Computers,2020,37(4):2505-2519.
.
[ 10] SHI Y Q, ZHAO C, PENG Z Q, et al. Analysis of the lag effect of embankment dam seepage based on delayed
mutual information[J] Engineering Geology,2018,234:132-137.
.
[ 11] ZHU Y T,ZHANG Z D,GU C S,et al. A coupled model for dam foundation seepage behavior monitoring and fore⁃
casting based on variational mode decomposition and improved temporal convolutional network[J] Structural Control
.
and Health Monitoring,2023,2023:3879096.
[ 12] 徐蔚,方春晖,王雪,等 . 考虑非线性环境量影响的混凝土重力坝坝基扬压力分析[J] 长江科学院院报,
.
2016,33(8):42-46,58.
.
[ 13] 庞琼,王士军,谷艳昌,等 . 基于滞后效应函数的土石坝渗流水位模型应用[J] 水土保持学报,2016,30
(2):225-229.
[ 14] 仇建春,蔡婷婷,蒋玮,等 . 大坝渗流统计模型在棉花滩水电站右岸绕坝渗流分析中的应用[J] 水电能源科
.
学,2013,31(5):49-51.
[ 15] WEI B W,GU M H,LI H K,et al. Modeling method for predicting seepage of RCC dams considering time-varying
and lag effect[J] Structural Control and Health Monitoring,2018,25(2):e2081.
.
[ 16] FANG W H, ZHANG W P, ZHANG C H, et al. Separation method of main and foreign water for the measuring
.
weirs of Danjiangkou earth-rock dam[J] Water,2022,14(22):3620.
[ 17] WANG S W,XU Y L,GU C S,et al. Monitoring models for base flow effect and daily variation of dam seepage ele⁃
ments considering time lag effect[J] Water Science and Engineering,2018,11(4):344-354.
.
[ 18] BEHFAR N, NOURANI V, SHARGHI E. Earthfill dam seepage analysis using ensemble artificial intelligence
based modeling[J] Journal of Hydroinformatics,2018,20(5):1071-1084.
.
[ 19] 余红玲,王晓玲,任炳昱,等 . 土石坝渗流性态分析的 IAO-XGBoost 集成学习模型与预测结果解释[J] 水利
.
学报,2023,54(10):1195-1209.
[ 20] ISHFAQUE M,DAI Q W,HAQ N U,et al. Use of recurrent neural network with long short-term memory for seep⁃
age prediction at Tarbela Dam,KP,Pakistan[J] Energies,2022,15 (9):123.
.
— 872 —

