ABSTRACT

This project presents a machine learning-powered web application designed to predict power quality issues based on key transmission line parameters. The application integrates a Flask-based backend, a MySQL database for user management, and a pre-trained Random Forest model to classify whether a power quality issue exists or not, If it exists it’ll classify the type of the power quality issue. By leveraging features such as voltage sag, harmonic distortion, temperature, and reactive power compensation, the model provides accurate, actionable insights. The system also incorporates user authentication and a simple, user-friendly interface, allowing operators and engineers to input real-time data and obtain quick predictions. This solution aims to improve decision-making processes in power quality management, reduce downtime, and enhance the overall reliability of the power distribution network.