Flood Prediction Based On Deep Learning Networks With Variational Mode Decomposition
Climate change increases the frequency of extreme weather events, causing river overflow floods that threaten human safety and ecosystems. Traditional flood prediction models face challenges due to fluctuations in water levels from topography and rainfall, leading to less accurate forecasts. This th...
| Main Author: | |
|---|---|
| Format: | Thesis |
| Language: | English |
| Published: |
2024
|
| Subjects: | |
| Online Access: | http://eprints.usm.my/62405/ http://eprints.usm.my/62405/1/NI%20CHENMIN%20-%20TESIS24.pdf |
| _version_ | 1848884975284256768 |
|---|---|
| author | Ni, Chenmin |
| author_facet | Ni, Chenmin |
| author_sort | Ni, Chenmin |
| building | USM Institutional Repository |
| collection | Online Access |
| description | Climate change increases the frequency of extreme weather events, causing river overflow floods that threaten human safety and ecosystems. Traditional flood prediction models face challenges due to fluctuations in water levels from topography and rainfall, leading to less accurate forecasts. This thesis aims to enhance flood prediction accuracy by developing and evaluating three new machine learning models that incorporate data decomposition, feature selection, and parameter optimization. The first two models use water level data for each hour. The first model utilizes hydrological data by integrating the Variational Mode Decomposition (VMD) method to reduce disturbances, along with Directional Bidirectional Long Short-Term Memory (BiLSTM) optimized with attention for forecasting purposes. The second model enhances prediction effectiveness by incorporating meteorological data specifically rainfall, humidity, and wind speed. This model emphasizes the benefits of VMD component classification and feature selection by considering water level changes to categorize Intrinsic Mode Functions (IMFs) obtained from the VMD method and using feature selection through the Pearson correlation method. The third model uses an optimized Gated Recurrent Unit - Temporal Convolutional Network (GRU-TCN) to forecast daily data at point estimates and confidence intervals. This model improves Kernel Density Estimate (KDE) predictions to assess forecast uncertainty more accurately and enhance model reliability. These three proposed models can overcome the weaknesses of traditional methods by utilizing real data from the Yangtze River station. |
| first_indexed | 2025-11-15T19:15:14Z |
| format | Thesis |
| id | usm-62405 |
| institution | Universiti Sains Malaysia |
| institution_category | Local University |
| language | English |
| last_indexed | 2025-11-15T19:15:14Z |
| publishDate | 2024 |
| recordtype | eprints |
| repository_type | Digital Repository |
| spelling | usm-624052025-06-03T04:22:28Z http://eprints.usm.my/62405/ Flood Prediction Based On Deep Learning Networks With Variational Mode Decomposition Ni, Chenmin QA75.5-76.95 Electronic computers. Computer science Climate change increases the frequency of extreme weather events, causing river overflow floods that threaten human safety and ecosystems. Traditional flood prediction models face challenges due to fluctuations in water levels from topography and rainfall, leading to less accurate forecasts. This thesis aims to enhance flood prediction accuracy by developing and evaluating three new machine learning models that incorporate data decomposition, feature selection, and parameter optimization. The first two models use water level data for each hour. The first model utilizes hydrological data by integrating the Variational Mode Decomposition (VMD) method to reduce disturbances, along with Directional Bidirectional Long Short-Term Memory (BiLSTM) optimized with attention for forecasting purposes. The second model enhances prediction effectiveness by incorporating meteorological data specifically rainfall, humidity, and wind speed. This model emphasizes the benefits of VMD component classification and feature selection by considering water level changes to categorize Intrinsic Mode Functions (IMFs) obtained from the VMD method and using feature selection through the Pearson correlation method. The third model uses an optimized Gated Recurrent Unit - Temporal Convolutional Network (GRU-TCN) to forecast daily data at point estimates and confidence intervals. This model improves Kernel Density Estimate (KDE) predictions to assess forecast uncertainty more accurately and enhance model reliability. These three proposed models can overcome the weaknesses of traditional methods by utilizing real data from the Yangtze River station. 2024-09 Thesis NonPeerReviewed application/pdf en http://eprints.usm.my/62405/1/NI%20CHENMIN%20-%20TESIS24.pdf Ni, Chenmin (2024) Flood Prediction Based On Deep Learning Networks With Variational Mode Decomposition. PhD thesis, Universiti Sains Malaysia. |
| spellingShingle | QA75.5-76.95 Electronic computers. Computer science Ni, Chenmin Flood Prediction Based On Deep Learning Networks With Variational Mode Decomposition |
| title | Flood Prediction Based On Deep Learning Networks With Variational Mode Decomposition |
| title_full | Flood Prediction Based On Deep Learning Networks With Variational Mode Decomposition |
| title_fullStr | Flood Prediction Based On Deep Learning Networks With Variational Mode Decomposition |
| title_full_unstemmed | Flood Prediction Based On Deep Learning Networks With Variational Mode Decomposition |
| title_short | Flood Prediction Based On Deep Learning Networks With Variational Mode Decomposition |
| title_sort | flood prediction based on deep learning networks with variational mode decomposition |
| topic | QA75.5-76.95 Electronic computers. Computer science |
| url | http://eprints.usm.my/62405/ http://eprints.usm.my/62405/1/NI%20CHENMIN%20-%20TESIS24.pdf |