A Machine Learning-Based Model with Graphical User Interface for Draft Prediction of Mouldboard Ploughing in Sandy Clay Loam Soil

A machine learning (ML)-based model was created to predict the draft force required for mouldboard ploughing in sandy clay loam soil. The ML model was trained using a data set comprising experimental measurements of draft requirements from ploughing operated at three distinct tillage depth levels and various speeds. Three prominent machine learning algorithms, namely Random Forests, Linear Regression, and Gradient Boosting, were employed to construct the ML model. Input parameters for the model included cone index, forward speed, and depth of operation, while the output parameter was the draft force. Among three algorithms tested, the Gradient Boosting demonstrated exceptional accuracy, with R2 value of 0.983 for the test data set. As a result, it was selected as the optimal ML model for draft force prediction. Further, to facilitate easy and practical use, a graphical user interface (GUI) was developed using Tkinter. With this GUI, users can effortlessly input the cone index, forward speed, and depth of operation to obtain predictions of draft force. The integration of developed ML model with the GUI provides significant potential for enhancing tillage operations in sandy clay loam soil by enabling more accurate draft prediction.