Landslide Displacement Prediction Model Based on Optimal Decomposition and Deep Attention Mechanism
Landslide displacement forecasting is crucial for disaster prevention and risk management, as it enables timely warnings and effective mitigation strategies. However, the highly nonlinear and complex nature of landslide displacement poses significant challenges for accurate prediction. To address th...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2025-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/10928339/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1850275832475418624 |
|---|---|
| author | Shuai Ren Kamarul Hawari Ghazali Yuanfa Ji Samra Urooj Khan |
| author_facet | Shuai Ren Kamarul Hawari Ghazali Yuanfa Ji Samra Urooj Khan |
| author_sort | Shuai Ren |
| collection | DOAJ |
| description | Landslide displacement forecasting is crucial for disaster prevention and risk management, as it enables timely warnings and effective mitigation strategies. However, the highly nonlinear and complex nature of landslide displacement poses significant challenges for accurate prediction. To address this, this study proposes an advanced forecasting framework integrating the Chebyshev Levy Flight-Sparrow Search Algorithm (CLF-SSA) with Variational Mode Decomposition (VMD) to enhance decomposition accuracy and optimize parameter selection. The trend component is modeled using the Autoregressive Integrated Moving Average (ARIMA) with a grid search strategy, while the periodic component is predicted using a Bidirectional Long Short-Term Memory network with an Attention mechanism (BiLSTM-Attention), which dynamically adjusts the contribution of influencing factors. Grey Relational Analysis (GRA) is further employed to identify key external driving factors, enhancing prediction accuracy. Experimental results demonstrate that the proposed model significantly improves predictive performance, reducing the Root Mean Square Error (RMSE) by 60% compared to the traditional XGBoost model and by 33% compared to the Empirical Mode Decomposition-BiLSTM (EMD-BiLSTM) model. Moreover, the Mean Absolute Scaled Error (MASE) analysis confirms the robustness of the model in capturing both short-term fluctuations and long-term trends. Given its superior predictive accuracy and practical applicability, this approach provides valuable technical support for landslide monitoring and early warning systems. |
| format | Article |
| id | doaj-art-67c92e51fa8a4a2bbe88bacf6097bea1 |
| institution | OA Journals |
| issn | 2169-3536 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj-art-67c92e51fa8a4a2bbe88bacf6097bea12025-08-20T01:50:33ZengIEEEIEEE Access2169-35362025-01-0113515735158810.1109/ACCESS.2025.355173010928339Landslide Displacement Prediction Model Based on Optimal Decomposition and Deep Attention MechanismShuai Ren0https://orcid.org/0009-0001-3288-4598Kamarul Hawari Ghazali1https://orcid.org/0009-0004-5737-6559Yuanfa Ji2https://orcid.org/0000-0001-8092-6679Samra Urooj Khan3Faculty of Electrical and Electronics Engineering Technology, Universiti Malaysia Pahang Al-Sultan Abdullah, Pekan, Pahang, MalaysiaFaculty of Electrical and Electronics Engineering Technology, Universiti Malaysia Pahang Al-Sultan Abdullah, Pekan, Pahang, MalaysiaFaculty of Electrical and Electronics Engineering Technology, Universiti Malaysia Pahang Al-Sultan Abdullah, Pekan, Pahang, MalaysiaFaculty of Electrical and Electronics Engineering Technology, Universiti Malaysia Pahang Al-Sultan Abdullah, Pekan, Pahang, MalaysiaLandslide displacement forecasting is crucial for disaster prevention and risk management, as it enables timely warnings and effective mitigation strategies. However, the highly nonlinear and complex nature of landslide displacement poses significant challenges for accurate prediction. To address this, this study proposes an advanced forecasting framework integrating the Chebyshev Levy Flight-Sparrow Search Algorithm (CLF-SSA) with Variational Mode Decomposition (VMD) to enhance decomposition accuracy and optimize parameter selection. The trend component is modeled using the Autoregressive Integrated Moving Average (ARIMA) with a grid search strategy, while the periodic component is predicted using a Bidirectional Long Short-Term Memory network with an Attention mechanism (BiLSTM-Attention), which dynamically adjusts the contribution of influencing factors. Grey Relational Analysis (GRA) is further employed to identify key external driving factors, enhancing prediction accuracy. Experimental results demonstrate that the proposed model significantly improves predictive performance, reducing the Root Mean Square Error (RMSE) by 60% compared to the traditional XGBoost model and by 33% compared to the Empirical Mode Decomposition-BiLSTM (EMD-BiLSTM) model. Moreover, the Mean Absolute Scaled Error (MASE) analysis confirms the robustness of the model in capturing both short-term fluctuations and long-term trends. Given its superior predictive accuracy and practical applicability, this approach provides valuable technical support for landslide monitoring and early warning systems.https://ieeexplore.ieee.org/document/10928339/Landslide displacement predictionintelligent optimization algorithmvariational mode decompositionbidirectional long short term memory networkattention mechanism |
| spellingShingle | Shuai Ren Kamarul Hawari Ghazali Yuanfa Ji Samra Urooj Khan Landslide Displacement Prediction Model Based on Optimal Decomposition and Deep Attention Mechanism IEEE Access Landslide displacement prediction intelligent optimization algorithm variational mode decomposition bidirectional long short term memory network attention mechanism |
| title | Landslide Displacement Prediction Model Based on Optimal Decomposition and Deep Attention Mechanism |
| title_full | Landslide Displacement Prediction Model Based on Optimal Decomposition and Deep Attention Mechanism |
| title_fullStr | Landslide Displacement Prediction Model Based on Optimal Decomposition and Deep Attention Mechanism |
| title_full_unstemmed | Landslide Displacement Prediction Model Based on Optimal Decomposition and Deep Attention Mechanism |
| title_short | Landslide Displacement Prediction Model Based on Optimal Decomposition and Deep Attention Mechanism |
| title_sort | landslide displacement prediction model based on optimal decomposition and deep attention mechanism |
| topic | Landslide displacement prediction intelligent optimization algorithm variational mode decomposition bidirectional long short term memory network attention mechanism |
| url | https://ieeexplore.ieee.org/document/10928339/ |
| work_keys_str_mv | AT shuairen landslidedisplacementpredictionmodelbasedonoptimaldecompositionanddeepattentionmechanism AT kamarulhawarighazali landslidedisplacementpredictionmodelbasedonoptimaldecompositionanddeepattentionmechanism AT yuanfaji landslidedisplacementpredictionmodelbasedonoptimaldecompositionanddeepattentionmechanism AT samrauroojkhan landslidedisplacementpredictionmodelbasedonoptimaldecompositionanddeepattentionmechanism |