A novel voltage control system based on deep neural networks for MicroGrids including communication delay as a complex and large-scale system

Microgrids play an important role in stabilizing the electrical grid and they are the best route to develop green and sustainable energy. Since microgrids are expanding rapidly, it is necessary to consider the related control issues including power quality, bidirectional power flow, voltage and frequency control, and stability analysis. One of the main measurement challenges is the communication delay. It means the delay in sending data from the sensor or measuring unit to the processing unit. The communication delay gets more important when the microgrid is widespread and complex. In this paper, a novel soft switching voltage control system is proposed to solve the voltage control problem of a widespread micro-grid while there are time-varying communication delays. The novel soft switching method is based on a static output feedback controller and deep neural networks. Another novelty of this paper is considering the 33-bus microgrid as a large-scale system that helps develop local and central controllers. The simulation's results show the effectiveness of a soft switching controller in the presence of dynamic time-varying communication delays. It means that while encountering static communication delays, the static output feedback controller without a soft switching method is sufficient in a large-scale microgrid.