Parameter identification of induction machines with consideration of rotor core loss

Parameter estimation accuracy plays an extremely critical role in the control performance of an induction motor (IM). Conventional IM parameter estimation can be performed by a dc-test, a no load test, and a blocked-rotor test. The dc-test determines the equivalent resistance of the stator winding. The no-load test provides behavior related to the magnetic core. The rotor resistance can be derived from the blocked-rotor test. According to the no-load test, the rotor parameter should be irrelevant to the stator power. However, significant mismatch in active power can be observed between simulation and experiment from a squirrel cage IM (SCIM) operating under super and sub-synchronous revolution speed. The same result can also be obtained from a doubly-fed IM (DFIM) in the case of rotor open-circuit test. To compromise the mismatched active power resulted from the conventional equivalent model due to the no-load test, a resistance representing the rotor core loss is considered. A novel procedure is proposed to determine the parameters of the IM. The first step of the proposed procedure is to determine the stator resistance by the dc-test. The second step is to find out the stator core parameters by a no-load test performing at synchronous rotor speed. The rotor parameters and the turn ratio between stator and rotor windings are finally obtained by the load test and the blocked-rotor test. Experimental and simulation results regarding to a SCIM and a DFIM are performed to differentiate the accuracy between the conventional and the proposed IM models.