Key points are not available for this paper at this time.
The present research addresses the crucial issues in intelligent power quality management inside smart grids by merging machine learning models with an Internet of Things (IoT) infrastructure. The research focuses on renewable energy sources, notably solar and wind, employing multiple sensors such as temperature, wind speed, and windmill rotation speed to obtain complete environmental data. A microcontroller records the sensor readings and related power production, generating a dataset of 20,223 entries over a month. Four machine learning models—the Recurrent Neural Network (RNN), Decision Tree (DT), Support Vector Machine (SVM), and Artificial Neural Network (ANN)—are evaluated on 20% of the dataset after being trained on 80% of it. Thesse trained ML model predicts the electricity that can be produced in future based on the information collected from the sensor. Also if the climatic condition is known then the electricity produced in future can be predicted in future. From the experimental result it is seen that the ANN emerges as the top-performing model with an accuracy of 93.5%, displaying its ability to predict electricity. SVM closely follows with 92.3%, while DT exhibits a balanced performance at 89.7%. The RNN, specialized in handling sequential data, achieves an accuracy of 86.22%. The findings underline the usefulness of the integrated system for forecasting electricity production based on real-time environmental conditions, supporting informed decision-making for smart grid stability. This work advances the field of intelligent power quality management and offers a roadmap for maximising the integration of renewable energy sources through the use of machine learning and Internet of Things technology. As the study seek for sustainable energy solutions, this research moves the sector ahead by giving practical insights for robust and efficient smart grid operations.
Rajendran et al. (Wed,) studied this question.