Coupling prediction model for long‐term displacements of arch dams based on long short‐term memory network

Liu, W.; Pan, Jianwen; Ren, Yisha; Wu, Zhigang; Wang, Jinting

doi:10.1002/stc.2548

Cited by 90 publications

(49 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Until now, the LSTM model has been applied to assess the safety of industrial facilities, such as tailings ponds, as well as the heating and cooling equipment (Li et al 2019;Wang et al 2019). In the field of civil engineering, this model is ever employed to predict the failure of bearings, seismic response of nonlinear structures, and displacement of dams (Gu et al 2018;Zhang et al 2019a, b;Liu et al 2020). Relatively high prediction accuracy was obtained in these studies.…”

Section: Lstmmentioning

confidence: 99%

“…Kaloop et al (2019) estimated the safety behavior of the Incheon large span bridge with the ARMA model and revealed that the bridge is safe under traffic loads. Liu et al (2020) regarded the dynamic coupled extreme stresses of bridges as time series data and applied the Bayesian probability recursive processes to successfully predict the value of stresses. However, currently, there are rare studies on using time series prediction techniques for estimating the response of bridges under dynamic loads in coastal environment, which is essential in terms of the hazard prevention for coastal bridges.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Application of time series prediction techniques for coastal bridge engineering

Wei

Yan

et al. 2021

ABEN

View full text Add to dashboard Cite

In this study, three machine learning techniques, the XGBoost (Extreme Gradient Boosting), LSTM (Long Short-Term Memory Networks), and ARIMA (Autoregressive Integrated Moving Average Model), are utilized to deal with the time series prediction tasks for coastal bridge engineering. The performance of these techniques is comparatively demonstrated in three typical cases, the wave-load-on-deck under regular waves, structural displacement under combined wind and wave loads, and wave height variation along with typhoon/hurricane approaching. To enhance the prediction accuracy, a typical data preprocessing method is adopted and an improved prediction framework for the LSTM model after the rolling forecast prediction is proposed. The obtained results show that: (a) When making a prediction on data featured with periodic regularity, both the XGBoost and ARIMA models perform well, and the XGBoost model can make predictions multi-step ahead, (b) The ARIMA model can predict just one step ahead based on aperiodic dataset with limited amplitude more accurately, while the XGBoost and LSTM models can predict multi-step ahead with appropriate data preprocessing, and (c) All the three models can predict the data tendency with model updating over time, but the prediction accuracy of the LSTM model is more favorable. The successful application of these three machine learning techniques can provide guidance to resolve engineering problems with time-history prediction requirements.

show abstract

Section: Lstmmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Application of time series prediction techniques for coastal bridge engineering

Wei

Yan

et al. 2021

ABEN

View full text Add to dashboard Cite

show abstract

“…The forget gate determines to what extent to forget the previous output results and selects the optimal time lag for the input sequence. The output gate determines what to output based on input and the memory of the block [33]. At time t, the input, forget, input modulation, and output gates can be formulated as i t , f t ,c t and o t , respectively.…”

Section: Stacked Long-short Term Memory Neural Networkmentioning

confidence: 99%

The Prediction of Dam Displacement Time Series Using STL, Extra-Trees, and Stacked LSTM Neural Network

Gong

Shu

et al. 2020

IEEE Access

View full text Add to dashboard Cite

Dam behavior prediction is a fundamental component of dam structural health monitoring. By comparing the predictions and the observations, anomalies can be detected, and then the remedial measures can be executed in time. As the most intuitive monitoring indicators, deformation is often used to evaluate dam structural health status. In this research, we propose a novel combined model for predicting the dam displacement time series. First, the seasonal-trend decomposition based on Loess (STL)method is utilized to decompose the dam displacement time series into seasonal, trend, and remainder components. Then the extremely randomized trees(extra-trees) model is used to predict seasonal components based on the causal models and influencing factors, whereas the stacked Long-Short Term Memory (LSTM)model is used to predict trend and remainder components based on the numerical models and historical observation data. Finally, the predicted results of the three components are aggregated to obtain the total predicted dam displacement. Seven state-of-the-art methods are introduced as benchmark methods to verify the effectiveness and feasibility of the proposed model. To quantitatively evaluate and compare the prediction results, three evaluation indicators, and a statistic test method are introduced. The experimental results show that the proposed model is the best-performing method compared with other benchmark methods both in prediction accuracy and stability. This indicates the proposed novel combined model STL-extra-trees-LSTM is a promising method for predicting displacement time series. INDEX TERMS Dam behavior prediction, time series decomposition, deep learning, ensemble learning.

show abstract

“…However, simple functions may not appropriately represent the relationships between influencing factors and dam displacement. To solve this problem, machine learning algorithms, such as support vector machine (SVM) (Su et al, 2015), extreme learning machine (ELM) (Kang et al, 2017), neural network (NN) (Mata, 2011), long short-term memory (LSTM) network (Liu et al, 2020), random forests (RF) (Dai et al, 2019), boosted regression tree (Salazar et al, 2016), and relevance vector machine (RVM) (Chen et al, 2020), have been shown to possess strong data mining abilities aimed at nonlinear implicit relations and have been employed to establish monitoring models. Meanwhile, based on the multi-scale characteristics, the long term measured displacement time series can be separated in the frequency domain into several components by the wavelet decomposition and empirical mode decomposition, and so forth, and according to the same frequency of effect component and its influencing factor, these components can be distinguished, by which the tendency component is usually defined as the time effect displacement (Correˆa et al, 2016;Fu et al, 2019;Su et al, 2018;Wang et al, 2018).…”

Section: Introductionmentioning

confidence: 99%

Hydraulic-seasonal-time-based state space model for displacement monitoring of high concrete dams

et al. 2021

Transactions of the Institute of Measurement and Control

View full text Add to dashboard Cite

Displacement is the most intuitive reflection of the comprehensive behavior of concrete dams, especially the time effect displacement, which is a key index for the evaluation of the structural behavior and health status of a dam in long-term service. The main purpose of this paper is to establish a state space model for separating causal components from the measured dam displacement. This approach is conducted by initially proposing two equations, which are the state and observation equations, and model parameters are then optimized by the Kalman filter algorithm. The state equation is derived according to the creep deformation of dam concrete and foundation rock and is used to preliminarily predict the dam time effect displacement. Considering the generally recognized three components of dam displacement, the hydraulic-seasonal-time (HST) model is used to establish the observation equation, which is used to update the time effect displacement. The efficiency and rationality of the established state space model is verified by an engineering example. The results show that the hydraulic component separated by the state space model only contains the instantaneous elastic hydraulic deformation, while the hysteretic elastic hydraulic deformation is divided into the time effect component. The inverted elastic modulus of dam body concrete is an instantaneous value for the state space model but a comprehensive reflection of the instantaneous and hysteretic elastic deformation ability for the HST model, where the hysteretic elastic deformation is a part of the hydraulic component. For the Xiaowan arch dam, the inverted values are 42.9 and 36.7 GPa for the state space model and HST model, respectively. The proposed state space model is useful to improve the interpretation ability of the separated displacement components of concrete dams.

show abstract

Coupling prediction model for long‐term displacements of arch dams based on long short‐term memory network

Cited by 90 publications

References 20 publications

Application of time series prediction techniques for coastal bridge engineering

Application of time series prediction techniques for coastal bridge engineering

The Prediction of Dam Displacement Time Series Using STL, Extra-Trees, and Stacked LSTM Neural Network

Hydraulic-seasonal-time-based state space model for displacement monitoring of high concrete dams

Contact Info

Product

Resources

About