Design and Implementation of Short-Term Load Forecasting using LightGBM

In the modern world, energy efficiency and smart systems are more necessary than ever before, making it imperative to possess a system capable of monitoring and forecasting power consumption in real-time. This research presents an advanced Data Acquisition System (DAS) combining hardware sensors, cloud computing, and machine learning to deliver accurate power monitoring and forecasting. The system employs sensors connected to an Arduino and ESP8266, which wirelessly transmit voltage and current information to Firebase for processing and storage. A machine learning algorithm implemented in Python subsequently forecasts power demand, with the outcomes displayed on a user-friendly web dashboard developed with Flask. This dashboard updates dynamically, displaying real-time power information and visualizing predictions every five minutes. Through the implementation of REST APIs, the system ensures smooth and efficient data transmission without requiring local storage. This research provides a real-world and cost-efficient solution that can integrate energy management systems with intelligent, real-time knowledge for better decision-making.