IoT based pump performance monitoring for agricultural irrigation and predictive maintenance with machine learning models

Irrigation pumps are used for agricultural irrigation in rural areas. Periodic maintenance of irrigation pumps in rural agricultural areas is costly and time-consuming; early detection of problems with real-time information avoids cost and time loss. The main objective of this study is to monitor the performance of the pump installed for agricultural irrigation using the Internet of Things (IoT) and to apply regression models to the system data and to provide the optimum parameter values for the models with genetic algorithm (GA). Firstly, the automation system was installed and the temperature, humidity, current, voltage and pump vibration data were transferred to an IoT platform the named "Thingspeak". Then, an application was created with the Qt user interface creation toolkit. Thanks to this application, malfunctions occurring in the system were instantly notified to the user. Predictive maintenance of the system consists of a two-stage process in the field of machine learning. In the first stage, four different method models were evaluated. These models are linear regression, polynomial regression (PR), random forest regression (RFR) and support vector regression (SVR). In the second stage, the optimal parameters of PR, SVR and RFR models were determined using GA. As a result, the random forest regression model was found to be the most appropriate model for the dataset. It is hoped that this thesis will contribute to researchers interested in IoT-based pump performance monitoring for agricultural irrigation and predictive maintenance estimation using machine learning models.