RANCANG BANGUN SISTEM PREDIKSI BANJIR BERBASIS INTERNET OF THINGS (IoT)

Floods are one of the most frequent natural disasters in Indonesia. This is due to several factors, including high rainfall, flood-prone geographical conditions, and lack of public awareness of the importance of disaster mitigation. One of the problems faced in flood disaster mitigation is the lack of accurate and timely information. Some of the methods are the most effective ways to detect the occurrence of Radian Basis Function (RBF), Support Vector Regression (SVR) and Long Short-Term Memory (LSTM) floods. Some of the basic components for making this prediction are river water level and rainfall which are carried out based on the Internet of Things. And these basic components are supported by several excellent features with the existence of the Artificial Intelligence system. In terms of system response, buzzers function well as alarms to provide warnings when conditions have the potential to flood. In predicting flood disasters, the 3 machine learning models used all showed an accuracy above 60%. This prediction system is perfect for providing early warning in flood-prone areas. With data from various sensors collected in real-time, preventive and evacuation measures can be taken to reduce the impact of disasters. To combine the advantages of the various models, further development can include merging with additional sources such as weather stations and updated machine learning algorithms.