基于改进组合深度学习模型的大坝位移预测研究  

Research on Dam Displacement Prediction Based on Improved Combined Deep Learning Model

作　　者：任传栋 王志真 刘淑萍 刘洪伟 侯龙潭 REN Chuan-dong;WANG Zhi-zhen;LIU Shu-ping;LIU Hong-wei;HOU Long-tan(Shandong Provincial Water Resources Survey and Design Institute Co.,Ltd.,Jinan 250013,China;Shandong Agricultural Exchange and Cooperation Center,Jinan 250013,China;Shandong Hydraulic Engineering Construction Quality and Safety Center,Jinan 250013,China)

机构地区：[1]山东省水利勘测设计院有限公司,山东济南250013 [2]山东省农业交流合作中心,山东济南250013 [3]山东省水利工程建设质量与安全中心,山东济南250013

出　　处：《水电能源科学》2023年第10期100-103,61,共5页Water Resources and Power

摘　　要：大坝位移可直接影响大坝的质量和运行安全,为找出大坝位移的合理预测模型,以时间卷积神经网络模型为基础(TCN),采用遗传算法对麻雀搜索算法(SSA)、灰狼算法(GWO)和蝙蝠算法(BA)三种仿生算法进行改进,得到MSSA、MGWO、MBA三种优化算法,并引入深度置信网络模型(DBN)构建了D-MSSA-TCN、D-MGWO-TCN、D-MBA-TCN三种组合赋权模型,以均方根误差、决定系数、平均绝对误差、效率系数和GPI指数为精度指标体系,结果表明在三种优化仿生算法中,MSSA算法的运行效率及精度最高,三种组合模型的精度显著高于其余模型,其中D-MSSA-TCN模型在所有模型中精度最高,可推荐用于估算坝体位移。Dam displacement can directly affect the quality and operation safety of the dam.To find out the prediction model of the dam displacement,the temporal convolutional neural network model was used to predict the dam displacement.Three bionic algorithms of the sparrow search algorithm(SSA),the gray wolf algorithm(GWO) and the bat algorithm(BA) were improved by genetic algorithm,and three optimization algorithms including MSSA,MGWO and MBA were obtained.Taking root mean square error,determination coefficient,mean absolute error,efficiency coefficient and GPI index as precision index system,three combined weighted models including D-MSSA-TCN,D-MGWO-TCN and D-MBA-TCN were constructed based on the deep belief network model(DBN).The results show that the MSSA algorithm had the highest operating efficiency and accuracy among all the algorithms.The accuracy of the three combined models was significantly higher than the rest of the models.The D-MSSA-TCN model had the highest accuracy among all models and can be recommended for estimating dam displacement.

关 键 词：坝体位移 时间卷积神经网络 麻雀搜索算法 遗传算法 深度置信网络模型 

分 类 号：TV698.1[水利工程—水利水电工程]