Implementation of deep learning-based controller for microgrid system for irrigation purpose

The hybrid wind and solar renewable involvement to meet the energy needs practically faces a lot of challenges. The main target to achieve is the power point tracking for both of the energy systems. In order to address the human-less irrigation process, this paper advances the deep learning-centered single-controller approach. The system is designed for the stand-alone energy system. So, the controller is responsible for achieving the peak power point for hybrid energy systems, battery, and load management. To carry out the said tasks, a deep learning-based MPPT controller is implemented and analyzed with the hardware setup. The federated learning algorithm is utilized as the controller, which by itself learns and teaches the controller to train at every moment for its effective learning and implementation process. The results show that the system performs well, with the tracking time for the MPPT achievement is 0.024 s with the designed algorithm by carrying out multiple tasks at a time.