A novel optimization-tuned hierarchical artificial neural network for managing renewable energy sources in smart grid

A smart grid that integrates additional renewable energy sources (RES) requires advanced management strategies to increase dependability and efficiency. Accurate estimations of energy consumption and RES generation are essential to the functioning of the smart grid. But the evaluation procedure is still difficult due to the unpredictable nature of RES and the diverse energy consumption patterns of users. The research introduced an innovative approach, the modified coot-optimized tuned hierarchical artificial neural network (MCO-HANN), for managing RES in smart grids. To use RES data, with continuous wavelet transform (CWT) applied for pre-processing to extract significant features from energy usage data. A python-based implementation of the MCO-HANN approach is evaluated using various metrics, including RMSE (0.0435), precision (96.2%), recall (95.4%), MAE (0.035), and accuracy (97.6%). The results of a comparative analysis show that the proposed MCO-HANN approach outperforms existing techniques in managing RES within smart grids. However, the model's functionality is influenced by the quality of input data, and its scalability can face challenges when applied to large, real-world smart grid environments. The method offers a viable way to maximize the incorporation of RES in upcoming smart grid systems.