Published Paper Details:

The proposed project is about creating a complete machine learning system that helps industries predict how much energy they will use and also check how efficiently a smart grid system is working. The main aim of the project is to manage energy in a smarter and more efficient way. It combines energy prediction and efficiency checking into one connected system. This helps industries reduce energy waste and improve their overall performance. The system uses modern technologies like machine learning and data analysis to make accurate predictions. It is designed to be practical and useful in real industrial environments, especially in industries like steel manufacturing where energy usage is usually very high. By predicting energy usage early, companies can plan their operations better and avoid sudden increases in electricity demand. This also helps reduce electricity costs and maintain stable power systems. In the first stage of the project, machine learning models are trained using a dataset collected from a steel industry. This dataset contains information about how machines use electricity during different operations. The system uses two important models called Random Forest and Long Short-Term Memory (LSTM). These models are good at finding patterns in complex data and understanding how different factors affect energy usage. For example, they analyze values such as load type, power factor, voltage, current, and other electrical measurements. These values change over time, and the models learn how these changes influence energy consumption. The Random Forest model works by combining many small decision trees to produce accurate predictions. It is useful for handling large amounts of data and identifying the most important factors that affect energy usage. The LSTM model is a type of deep learning model that is specially designed to understand time-based data. It can remember past information and use it to predict future values. This makes it very suitable for predicting energy consumption because energy usage usually follows patterns over time. Using both models together improves prediction accuracy. After training the models, the system predicts short-term energy consumption in units called kilowatt-hours (kWh). These predictions show how much energy the industry will likely use in the near future. The predicted energy value is then added to another dataset called a synthetic smart grid dataset, which represents how electricity flows through a smart grid system. A smart grid is a modern electricity network that uses digital technology to monitor and control energy distribution. In the next step, the system calculates an Energy Efficiency Score based on the predicted energy consumption. This score shows how efficiently the grid is using energy. A higher score means better efficiency, while a lower score means energy is being wasted. The system uses fixed threshold values to classify the performance as either efficient or inefficient. This makes it easy for users to quickly understand the condition of the grid. Another machine learning model called a Random Forest Classifier is then used to predict grid efficiency using factors such as power demand, renewable energy contribution, weather conditions, temperature, and environmental data. The model analyzes these factors and predicts whether the system will operate efficiently. It also shows which factors are most important in affecting energy efficiency, helping engineers make better decisions to improve performance. Finally, the entire system is connected to a web application . Users can view results through dashboards and charts in a simple and user-friendly interface. Overall, this project provides a smart and scalable solution for energy management by helping industries save energy, reduce costs, improve system stability, and support sustainable energy practices.