Development of particle swarm optimization based rainfall-runoff prediction model for Pahang River, Pekan
Flooding is a natural disaster which has been occurring annually throughout the whole world. The disaster, such as other natural catastrophe could only be mitigated rather than it being completely solved. Runoff prediction proved to be very vital in pre-flooding management system. In recent years, A...
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Conference or Workshop Item |
| Language: | English English |
| Published: |
The Institute of Electrical and Electronics Engineers, Inc.
2016
|
| Subjects: | |
| Online Access: | http://irep.iium.edu.my/58177/1/58177_Development%20of%20Particle%20Swarm.pdf http://irep.iium.edu.my/58177/2/58177_Development%20of%20Particle%20Swarm_SCOPUS.pdf http://irep.iium.edu.my/58177/ http://ieeexplore.ieee.org/document/7808330/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Institution: | Universiti Islam Antarabangsa Malaysia |
| Language: | English English |
| Summary: | Flooding is a natural disaster which has been occurring annually throughout the whole world. The disaster, such as other natural catastrophe could only be mitigated rather than it being completely solved. Runoff prediction proved to be very vital in pre-flooding management system. In recent years, Artificial Neural Network has been applied in various prediction models of hydrological system. It is proposed to model the rainfall-runoff system of Pahang River in Pekan. Mean rainfall data of 5 hydrological stations are used as the input and water level data as the output. The Artificial Neural Networks are trained with Particle Swarm Optimization. The performances of Artificial Neural Networks were measured with Ackley cost function value. Neural network configuration of 450 number of maximum iteration, 6 number of particles and 1.9 and 2.0 values of Particle Swarm Optimization parameter constant for global best (c1) and Particle Swarm Optimization constant for personal best (c2) respectively shows the highest global best function value. The neural network configuration of 300 number of maximum iteration, 3 numbers of particles and 2.2 value of (c1) and (c2) produces lowest global best function value. The output shows Artificial Neural Network trained by Particle Swarm Optimization can successfully model rainfall-runoff. © 2016 IEEE. |
|---|