DNN-Driven Voltage Regulation in Wind Energy Systems with Electric Spring

An Electric Spring (ES) based wind energy system is proposed in this paper for robust voltage management using a Deep Neural Network (DNN). In this research we look at the outline and implementation of a DNN that controls the ES in a wind energy system with difficult loading conditions. Although Electric spring is used for voltage control before, the main improvement is the robustness that the well-trained DNN brings. To train the DNN parameters, a typical distribution network was used to generate the dataset. The learned model was then evaluated on unseen scenarios. By using fewer sensors and being less reliant on grid variables, the results shows that the DNN-driven ES performs better than previous ES implementations.