Operational Performance Assessment of PV-Powered Street Lighting A Comparative Study of Different Machine Learning Prediction Models

This project aims to assess the operational performance of photovoltaic (PV)-powered street lighting systems using machine learning prediction models. The study focuses on comparing the performance of various algorithms, including Hybrid Models (Linear Regression and Random Forest), CatBoost Regressor, and XGBoost Regressor, to predict the efficiency and energy consumption of PV-powered street lighting. By utilizing environmental and operational data such as traffic density, ambient light, weather conditions, power state, and geographical location, the system evaluates the sustainability and efficiency of these lighting systems. Data is uploaded via a user-friendly interface, and the models are evaluated on performance metrics such as Mean Absolute Error (MAE), Mean Squared Error (MSE), Root Mean Squared Error (RMSE), and R² Score. The predictive capabilities of the models are used to provide actionable insights for optimizing energy consumption, helping to improve the sustainability of street lighting systems. The project uses Flask as the web framework, with joblib for model deployment and MySQL for database management. The end goal is to enable more efficient and environmentally friendly street lighting solutions by leveraging machine learning for performance assessment and prediction.

Keywords:

Photovoltaic, Street Lighting, Machine Learning, Prediction Models, Hybrid Model, CatBoost, XGBoost, Energy Efficiency, Sustainability, Flask, Model Evaluation, Performance Metrics.