Artificial Neural Network Based Identification of Multi-Operating-Point Impedance Model
Mengfan Zhang, Xiongfei Wang, Dongsheng Yang, Mads Græsbøll Christensen
Abstract
The black-box impedance model of voltage source inverters (VSIs) can be measured at their terminals without access to internal control details, which greatly facilitate the analysis of inverter-grid interactions. However, the impedance model of VSI is dependent on its operating point and can have different profiles when the operating point is changed. This letter proposes a method for identifying the impedance model of VSI under a wide range of operating points. The approach is based on the artificial neural network (ANN), where a general framework for applying the ANN to identify the VSI impedance is established. The effectiveness of the ANN-based method is validated with the analytical impedance models.