Multi-Layered Artificial Neural Network Flood Prediction System with Rain Gauge, Temperature Humidity Pressure Sensor, Ultrasonic Sensor, Soil Moisture Sensor and Anemometer

Abstract

Located near the Pacific Ocean, Philippines is along in one of the places regarded as a Typhoon Belt having an average of 19 to 20 tropical cyclones occurring every year. These tropical cyclones leave a devastating effect especially to properties, infrastructures and lives of the people. The development of a system that can predict the flood level based on different weather parameters such as ambient temperature, relative humidity, barometric pressure and wind speed integrated with rain gauge, soil moisture sensor and water level sensor were conducted in this study. MATLAB Neural Network Tool was used to develop the Multi-Layer Artificial Neural Network prediction model whose inputs were based on different weather data values from the sensors and targets the water level data as the output. Upon development of the model using MATLAB, the training, test and validation datasets which was divided into 70%, 15% and 15% of the data shows a goodness- of-fit of 0.93868, 0.94381 and 0.95607 respectively. The weights and biases of the prediction model were then integrated to the system and was tested yielding to an RMSD value of 5.6349.