地理空间信息2024，Vol.22Issue(1)：92-95,4.DOI:10.3969/j.issn.1672-4623.2024.01.022
改进水循环优化BP神经网络的大坝变形预测
Dam Deformation Prediction Method Based on IWCA-BP Neural Network
摘要
Abstract
The traditional dam deformation prediction method based on back propagation(BP)neural network has the problems of low predic-tion accuracy and poor noise robustness.We proposed an improved water cycle algorithm(IWCA)optimized BP neural network(IWCA-BP)model to realize high-precision prediction of building deformation trend.Firstly,we used empirical mode decomposition(EMD)to decompose high-dimensional complex deformation data into a series of intrinsic mode function(IMF)and the sum of residual terms.Then,we used the IWCA-BP neural network to model and predict each IMF separately.As a heuristic optimization algorithm,IWCA could quickly and accurately realize the global optimization of initial value of BP neural network,and improve the prediction performance.Finally,we used the actual defor-mation data of a concrete dam to carry out experiment.The results show that compared with Kalman filter,support vector machine,BP neural net-work and the latest particle swarm optimized random forest method,the proposed EMD-IWCA-BP method can achieve higher prediction accura-cy and noise robustness,and has a higher application prospect.关键词
变形预测/BP神经网络/EMD/噪声稳健性/水循环算法Key words
deformation prediction/BP neural network/EMD/noise robustness/water cycle algorithm分类
天文与地球科学引用本文复制引用
胡振东,郭明强..改进水循环优化BP神经网络的大坝变形预测[J].地理空间信息,2024,22(1):92-95,4.基金项目
国家自然科学基金资助项目(41971356、41701446) （41971356、41701446）
自然资源部城市国土资源监测与仿真重点实验室开放基金资助项目(KF-2020-05-011). （KF-2020-05-011）
