7 hours flood prediction modeling using NNARX structure: Case study Kedah
| Autor: | Abd Manan Samad, Ramli Adnan, Fazlina Ahmat Ruslan, Zainazlan Md Zain |
|---|---|
| Rok vydání: | 2014 |
| Předmět: |
Upstream (petroleum industry)
Artificial neural network Meteorology Flood myth business.industry Machine learning computer.software_genre Training (civil) Water level Autoregressive model Environmental science Artificial intelligence business MATLAB Downstream (networking) computer computer.programming_language |
| Zdroj: | ICCSCE |
| DOI: | 10.1109/iccsce.2014.7072758 |
| Popis: | Most of the countries around the world have paid great attention to flood water level prediction system because flood events may damage on people's life and property. However, since flood water level fluctuates highly nonlinear, it is a very difficult task to predict flood water level accurately. Since Artificial Neural Network is an effective technique for handling nonlinear problems, thus, this paper proposed a 7 hours ahead flood water level prediction modelling using Neural Network Autoregressive with Exogenous Input (NNARX) for flood prone area located in Kedah, Malaysia as case study. The model was developed using four inputs and one output. Three inputs were upstream stations water level and one input from water level differences at downstream flood location. The output was the predicted water level at downstream station. Simulation was done using Matlab Neural Network Toolbox. Results showNNARX modelling was able to predict flood water level ahead of time. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|