Daily rainfall runoff modeling using artificial neural network for sungai Sarawak Kanan upper catchment
This thesis reports the results predicted from artificial neural network (ANN) models for daily rainfall-runoff simulation, at Sungai Sarawak Kanan upper catchment. The system is monitored by four rainfall gauging stations namely Kampung Monggak, Krokong, Bau and Kampung Opar located upstream of th...
Saved in:
| Main Author: | |
|---|---|
| Format: | Final Year Project Report / IMRAD |
| Language: | en en |
| Published: |
Universiti Malaysia Sarawak (UNIMAS)
2005
|
| Subjects: | |
| Online Access: | http://ir.unimas.my/id/eprint/23717/1/Daily%20rainfall%20runoff%20modeling%20using%20artificial%20neural...%2824%20pgs%29.pdf http://ir.unimas.my/id/eprint/23717/4/Ong%20Khin%20Kiat.pdf http://ir.unimas.my/id/eprint/23717/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | This thesis reports the results predicted from artificial neural network (ANN) models for daily rainfall-runoff simulation, at Sungai Sarawak Kanan upper catchment.
The system is monitored by four rainfall gauging stations namely Kampung Monggak, Krokong, Bau and Kampung Opar located upstream of the river system and one river
stage gauging station namely Buan Bidi. Backpropagation network (BP) of multilayer perceptron (MLP) is used for daily runoff simulation. Input variables used are current
rainfall, antecedent rainfall and antecedent runoff while the output is current runoff. Several networks were trained and tested using data obtained from Department of
Irrigation and Drainage (DID) Sarawak. The effects of different types of training algorithms, different numbers of hidden neurons, different numbers of antecedent data
and different numbers of hidden layers were investigated to find the optimal neural network. Judging on coefficient of correlation R, one layered training algorithm trainoss (R = 0.839) with 150 hidden neurons and 5 days backdated performed the best for the
simulations. Therefore this make this study useful for heavy rainfall predictions and thus a good tool for flood warning. |
|---|