Fuzzy Logic-based Optimization of Flood Control System

Floods wreak havoc on the lives of people, making them a natural calamity of great significance. Flood catastrophes can have negative effects, including supplies and infrastructural destruction as well as fatalities. Urban floods can pose a major risk to public wellness, destroy property, and result in financial losses. To this level, it is imperative to improve flood control strategies through the installation of suitable facilities including detention reservoirs. Although smart systems may be used to regulate floods using detention reservoirs, there is a lack of research on the optimal design of such facilities in the face of flood threats. This paper presents a fuzzy logic-based control system for assessing the efficiency of flood drainage structures using three factors: hydrological efficiency, technological circumstance, and functional circumstance. The approach allows for a uniform evaluation of system dependability by synthesizing expertise information into 125 inference criteria. The system has been verified against experimental databases from drainage systems in Europe, and it performs well in a variety of functional and geographic scenarios. The platform has been coupled with Telegram and constructed with multiple detectors. Since the output needed for this research is in the type of values or linear equations, the Sugeno technique is more suited to be used in this investigation. Water stream and ultrasonic detectors are employed. The water stream detector determines the stream of water reaching the trial site at intervals of 0–11 liters per minute, while the ultrasonic detector measures the water concentration over a spectrum of 0–51 cm. With the help of the WiFi component and the NodeMCU ESP8266, statistics are transmitted to Telegram in real-time via the Firebase dataset. Water degree and release statistics are analyzed and Sugeno fuzzy logic is used for analyzing the outcomes. The mean deviation of the value from the ultrasonic detector, according to the investigation's findings, is 2.44%, or 97.59%. The mean inaccuracy displayed by the water stream detector is 0.207 liters per minute or an accuracy ratio of 87.07%.