Load Forecasting using Machine Learning Model

This paper presents a comprehensive system for load forecasting, a critical component in energy management. The system employs advanced machine learning techniques, specifically the Long Short-Term Memory (LSTM) model, to provide accurate and tailored load forecasts. The paper follows a streamlined flow, integrating data collection, pre-processing, model training, and user interaction. The system’s architecture involves a react front-end for user interaction and a flask backend for seamless communication with the machine learning model. Visual representations, such as line graphs, enhance the understanding of load patterns, and interpolation/extrapolation techniques contribute to the system’s forecasting accuracy. Experiments were involved using XGBoost model. Compared to XGBoost model and LSTM has better accuacy.