Subject Area: CONTROL SYSTEM ENGINEERING
This paper presents the modeling of a real-time flood detection and control system using machine learning techniques. The experimental and simulation methodology was employed to achieve the objective of this work. The study characterized an existing flood detection system and identified the technical challenges after data collection and analysis, then a nonlinear flood model was developed, and a sensor was designed to acquire real-time flood data from the environment, considering volume and pressure an sensing elements. A nonlinear model predictive control system was then modeled using flood data, artificial neural network and implemented in Simulink, which utilized previous flood behavior to forecast the future response of the system. The simulation results demonstrated that the new flood detection system achieved a regression of 1 after several iterations, indicating a good fit between the model and the actual flood data. Overall, the results of this study indicate that the proposed flood detection and control system has the potential to effectively detect and mitigate floods in real-time, thus reducing the loss of life and property caused by flooding.