SYNCHRONOUS MACHINE PARAMETER-IDENTIFICATION BY A MULTITIME SCALE TECHNIQUE

We propose a new method for the synchronous machine parameter identifi cation. We use a multitime scale decomposition to obtain reduced opera tional reactances. The identification method is based on the frequency responses. The synchronous machine is at a standstill and supplied by voltage pseudorandom binary sequences (PRBS). The output data are the currents in the windings. All the Park's model parameters of an alter nator of 13.5 kVa / 19.5 A / 50 Hz are estimated by an autoregressive with external input (ARX) model structure.