Δημιουργία συσκευής ανίχνευσης φραγής φρεατίου ομβρίων υδάτων σε συνδυασμό με την θεωρητική μελέτη προβλεπτικών αλγόριθμων για την πρόβλεψη πλημμύρας

Abstract

This thesis is divided into two parts. The first part refers to the implementation of a device, which enters a rainwater well/manhole and aims to detect its blockage. The device consists of an Arduino UNO microcontroller with a wi-fi module, in order for it to be converted into an IoT (Internet of Things) device and in combination with a float sensor that detects the blockage of the pipe inside the well, a water level sensor to measure its level water and humidity sensor that measures the humidity levels inside the well, we can determine the condition of a rainwater well/manhole. All data collected by the sensors is stored in the Cloud. In addition, a website has been set up, where it is updated with the current data of the sensors. The second part focuses on the theoretical study of various machine learning algorithms in order to achieve the prediction of a flood. With the data provided by the sensors in combination with other data such as the geomorphological data of an area, the millimeters and hours of rain, an algorithm can be implemented that will return the possibility of flooding depending on the condition of the wells. This work could help municipalities because they will be informed in real-time about the condition of the wells and will be able to act faster in case of need. It can also provide important information that will help in the reconstruction of defective wells.