AI BASED POWER CONSUMPTION ANALYSIS AND PREDICTION FOR ELECTRICAL APPLIANCES.

Abstract

Abstract —  Electricity consumption has been a subject of extensive research in the field of computer architecture for many years. The primary aim of this study is to provide useful guidelines to the machine learning community, enabling them to use and develop energy estimation techniques for machine learning algorithms. Various ensemble models, such as Linear Regression, Random Forest Regression, and LSTM, have been utilized to predict electricity consumption and achieve accurate results. Additionally, this project introduces the latest software tools that support electricity estimation principles, along with two use cases that reinforce the investigation of energy usage in machine learning. Ultimately, this study predicts future energy usage, which is instrumental in enabling the grid to accurately provide energy by leveraging smart meters that provide insights into appliance usage patterns. This information helps users determine when they require more or less energy, making it immensely valuable.