ABSTRACT
The research explores an IoT system for forecasting energy consumption in distribution stations using smart sensors and machine learning models. Real-time data collection and experiments demonstrate strong correlations between sensor readings and power consumption, with the ANN model achieving high accuracy. Naive Bayes, Random Forest, and Decision Trees also show competitive results. The study evaluates the models’ comprehensibility, computational efficiency, and adaptability for practical use. Integrating IoT and machine learning improves real-time monitoring and predictive analytics, optimizing resource allocation in smart grid environments. These findings lay the groundwork for more efficient and sustainable energy management systems.
