A Dam Deformation Prediction Model Based on ARIMA-LSTM

Xu, Guoyan; Jing, Zixu; Mao, Yingchi; Su, Xinyue

doi:10.1109/bigdataservice49289.2020.00040

Cited by 9 publications

(6 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Step 2 for the model evaluation. The XGBoost model feasibility was evaluated by its comparison with the model based on the LightGBM algorithm (Xu et al, 2021). The respective flowchart is depicted in Figure 7 (Su et al, 2016;Xu et al, 2021).…”

Section: Xgboost-based Prediction Proceduresmentioning

confidence: 99%

“…The XGBoost model feasibility was evaluated by its comparison with the model based on the LightGBM algorithm (Xu et al, 2021). The respective flowchart is depicted in Figure 7 (Su et al, 2016;Xu et al, 2021). Four kinds of ML regression algorithms, namely the support vector machine (SVM), Random Forest, LightGBM, and XGBoost, were realized, yielding the pipeline bottom stress regression values.…”

Section: Xgboost-based Prediction Proceduresmentioning

confidence: 99%

“…In recent years, machine learning (ML), which could comprehensively learn potential correlations of input and output variables, has been widely applied to nonlinear multiparameter regression problems (Prayogo et al, 2020). Thus, Xu et al (2021) combined the XGBoost (an optimized distributed gradient boosting library designed to be highly efficient, flexible, and portable), backpropagation (BP) neural network, and linear regression method to establish a dam deformation prediction model. They concluded that it was problematic to accurately predict the peak of training samples by the tree algorithm.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Pipeline Stress Test Simulation Under Freeze-Thaw Cycling via the XGBoost-Based Prediction Model

Teng

et al. 2022

Front. Earth Sci.

View full text Add to dashboard Cite

This study conducted ten freeze-thaw cyclic tests to clarify the effect of freeze-thaw cycles on the forces acting on the buried oil pipeline. The stress evolution in the Q345 steel pipeline versus the number of freeze-thaw cycles was obtained. The test results were consistent with the COMSOL simulation of the effect of different moisture contents on the pipeline bottom stress. Besides the proposed XGBoost model, eleven machine-learning stress prediction models were also applied to 10–20 freeze-thaw cycling tests. The results showed that during the freeze-thaw process, the compressive stress at the pipeline bottom did not exceed −69.785 MPa. After eight freeze-thaw cycles, the extreme value of the principal stress of -252.437MPa, i.e., 73.17% of the yield stress, was reached. When the initial moisture content exceeded 20%, the eighth freeze-thaw cycle’s pipeline stress decreased remarkably. The XGBoost model effectively predicted the pipeline’s principal stress in each cycle of 10 freeze-thaw cyclic tests, with R2 = 0.978, MSE = 0.021, and MAE = 0.102. The above compressive stress fluctuated from −131.226 to −224.105 MPa. The predicted values well matched the experimental ones, being in concert with the “ratcheting effect” predicted by the freeze-thaw cycle theory. The results obtained provide references for the design, operation, and maintenance of buried oil pipelines.

show abstract

Section: Xgboost-based Prediction Proceduresmentioning

confidence: 99%

Section: Xgboost-based Prediction Proceduresmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Pipeline Stress Test Simulation Under Freeze-Thaw Cycling via the XGBoost-Based Prediction Model

Teng

et al. 2022

Front. Earth Sci.

View full text Add to dashboard Cite

show abstract

“…Han et al [15] established the SA-RELM model based on time series, achieving improved accuracy in ground subsidence prediction for excavation. Xu et al [16] constructed an ARIMA-LSTM model to predict nonlinear feature data in dam deformations. Kim et al [17,18] employed machine learning algorithms to effectively predict tunnel surface settlement, enhancing the prediction capabilities for surface settlement in urban tunnel construction sites under complex excavation conditions.…”

Section: Introductionmentioning

confidence: 99%

Mine Surface Settlement Prediction Based on Optimized VMD and Multi-Model Combination

Shen,

2023

Processes

View full text Add to dashboard Cite

The accurate prediction of mining area surface deformation is essential to preventing large-scale coal mining-related surface collapse and ensure safety and daily life continuity. Monitoring subsidence in mining areas is challenged by environmental interference, causing data noise. This paper employs the Sparrow Search Algorithm, which integrates Sine Cosine and Cauchy mutation (SCSSA), to optimize variational mode decomposition (VMD) and combine multi-models for prediction. Firstly, SCSSA is employed to adaptively determine the parameters of VMD using envelope entropy as the fitness value. Subsequently, the VMD method optimized using SCSSA adaptively decomposes the original mining area subsidence data sequence into various sub-sequences. Then, SCSSA-VMD is applied to adaptively decompose the original mining subsidence data sequence into multiple sub-sequences. Meanwhile, using sample entropy, the sub-sequences are categorized into trend sequences and fluctuation sequences, and different models are employed to predict sub-sequences at different frequencies. Finally, the prediction results from different sub-sequences are integrated to obtain the final prediction of mining area subsidence. To validate the predictive performance of the established model, experiments are conducted using GNSS monitoring data from the 110801 working face of Banji Coal Mine in Bozhou. The results demonstrate the following: (1) The hybrid model enhanced the prediction accuracy and trends by decomposing the data and optimizing the parameters with VMD. It outperformed single models, reducing errors and improving predictive trends. (2) The hybrid model significantly improved the prediction accuracy for subsidence data at work surface monitoring stations. It is particularly effective at critical subsidence points, making it a valuable reference for safety in mining operations.

show abstract

“…For long and short-term memory model (LSTM) prediction, it has good fitting ability and can make up for the low prediction accuracy of the BP neural network. Compared with the single model, the nonlinear combined dynamic propagation rate model has Higher prediction accuracy and robustness [4]; Infectious disease models often have problems such as imperfect construction and unrealistic assumptions, so the combination of models is often applied to improve predictions, such as the combination of the SEIR and ARIMA models [5][6][7], but there is strong uncertainty. We improved the ARIMA model, which is simple and has strong adaptability and good explanatory power.…”

Section: Model Introduction 21 Arimamentioning

confidence: 99%

Prediction of COVID-19 Data Using an ARIMA-LSTM Hybrid Forecast Model

et al. 2022

View full text Add to dashboard Cite

The purpose of this study is to study the spread of COVID-19, establish a predictive model, and provide guidance for its prevention and control. Considering the high complexity of epidemic data, we adopted an ARIMA-LSTM combined model to describe and predict future transmission. A new method of the ARIMA-LSTM model paralleling by weight of regression coefficient was proposed. Then, we used the ARIMA-LSTM model paralleling by weight of regression coefficient, ARIMA model, and ARIMA-LSTM series model to predict the epidemic data in China, and we found that the ARIMA-LSTM model paralleling by weight of regression coefficient had the best prediction accuracy. In the ARIMA-LSTM model paralleling by weight of regression coefficient, MSE = 4049.913, RMSE = 63.639, MAPE = 0.205, R2 = 0.837, MAE = 44.320. In order to verify the effectiveness of the ARIMA-LSTM model paralleling by weight of regression coefficient, we compared the ARIMA-LSTM model paralleling by weight of regression coefficient with the SVR model and found that ARIMA-LSTM model paralleling by weight of regression coefficient has better prediction accuracy. It was further verified with the epidemic data of India and found that the prediction accuracy of the ARIMA-LSTM model paralleling by weight of regression coefficient was still higher than that of the SVR model. In the ARIMA-LSTM model paralleling by weight of regression coefficient, MSE = 744,904.6, RMSE = 863.079, MAPE = 0.107, R2 = 0.983, MAE = 580.348. Finally, we used the ARIMA-LSTM model paralleling by weight of regression coefficient to predict the future epidemic situation in China. We found that in the next 60 days, the epidemic situation in China will become a steady downward trend.

show abstract

A Dam Deformation Prediction Model Based on ARIMA-LSTM

Cited by 9 publications

References 7 publications

Pipeline Stress Test Simulation Under Freeze-Thaw Cycling via the XGBoost-Based Prediction Model

Pipeline Stress Test Simulation Under Freeze-Thaw Cycling via the XGBoost-Based Prediction Model

Mine Surface Settlement Prediction Based on Optimized VMD and Multi-Model Combination

Prediction of COVID-19 Data Using an ARIMA-LSTM Hybrid Forecast Model

Contact Info

Product

Resources

About