Realtime Estimation of the Parameters and Fluxes of Induction Motors

The paper presents a new method for the realtime estimation of the parameters and fluxes of induction motors. The procedure is potentially useful for the design of self-commissioning drives, i.e., drives that can adjust controller parameters automatically for a wide range of motors and loads. Another possible application is for the detection of failures. In their recursive form, the algorithms can be used for adaptation to parameters that vary with time, and for estimation of the rotor fluxes in a field-oriented controlled drive. The estimation method is based on a standard model of the induction motor, expressed in rotor coordinates. It is assumed that current and position (or velocity) measurements are available. The rotor fluxes are not assumed to be measured. The interesting features of the method are that: (1) it does not rely on special tests such as the locked rotor test or the no-load test (instead, a broad range of motor responses can be used), (2) the method provides estimates of the rotor fluxes together with the estimates of the parameters, (3) measures of the uncertainties in the estimation procedure are given (this provides feedback as to the precision of the parameter estimates, as well as some guidance to choose excitation and to optimize the quality of the parameter estimation). Results for both simulated and experimental data are reported, and the ability of the algorithm to rapidly estimate the motor parameters is demonstrated. The experimental results are validated using different experiments and compared with those obtained using a standard frequency response test at standstill.