Identification of Self and Mutual Inductances of Electrically Excited Synchronous Machines Using Signals in Current Controller Without Extra Observer

Electrically excited synchronous machines (EESMs) have become a promising candidate for electric vehicles due to their high power density without using rare-earth materials. Accurate parameters are essential to achieve high-performance dynamic control. However, to precisely identify the parameters of an EESM is more challenging than other machines, since the number of parameters is larger and there are also couplings between the parameters. In this study, a method to identify the apparent and incremental values of self and mutual inductances is proposed. In this method, signals from the current controller are extracted and processed to eliminate the parameter errors, without building any extra observer. The method has been verified in both real-time simulations and experiments. It is shown that parameters are calibrated successfully. By using the identified parameters in the current controller, a fast and smooth current dynamic response is achieved.