High Impact Factor : 4.396 icon | Submit Manuscript Online icon |

Smart Flood Prediction System

Author(s):

Aneena Joy , IES College of Engineering, Chittilappilly, Thrissur; Dr. G Kiruthiga, IES College of Engineering, Chittilappilly, Thrissur

Keywords:

Flood, Hazard, Flood Prediction Model, NNARX, EKF

Abstract

Flood occurrences are the most hated environmental hazard around the world. This is due to floods threaten human life and furthermore affect the economy of the involved country. Floods are usually due to the season monsoon and heavy rains causing flash floods usually in urban area. Therefore, it is a must for researchers around the world to find a solution for this problem. A reliable and practical flood prediction model is in need to predict flood occurrence ahead of time. Thus, this paper proposes the modelling of flood prediction system using Extended NNARX (Neural Network Autoregressive with Exogenous Input) with EKF (Extended Kalman Filter). Various environmental parameters required for flood prediction such as temperature, humidity, water level, water quality are obtained by using various sensors. The data continuously obtained from these sensors are used to train the Extended NNARX model. The trained Extended NNARX model can predict the occurrence and non-occurrence of flood by taking the current climatic conditions as input and comparing it with the previous values. Weather prediction is also possible from the sensor readings. Alert notifications are sent to the authorities as well as the public whenever occurrence of flood is predicted. Public is also alerted through a buzzer. The sensor values and its graphical representation are displayed through a LCD and a monitor respectively that are placed at public locations so that the public will be aware of the current climatic conditions of their location in real time.

Other Details

Paper ID: IJSRDV7I40704
Published in: Volume : 7, Issue : 4
Publication Date: 01/07/2019
Page(s): 701-704

Article Preview

Download Article