Identification ofelectrical parameters androtor speed of induction motorusing radial basis neural network
C. N. R. S-Supelec · 2004
A technique forparameters androtorspeed of induction motorusingthecombination ofhigh-gain observer andradial basis function neuronal predictor istreated inthis paper. Thealgorithms developed herearepotentially useful for thedesign ofthedrives thatcanadjust controller parameters automatically. Another possible application isforthedetection offailure. Inthefirst scheme, leakage coefficient androtortime- constant areestimated using themeasured rotorspeed. stator current andvoltage. Inthesecond scheme, rotorresistance is estimated whentherotor speed isavailable while inthethird scheme, bothrotor resistance androtor speed areidentified. All theparameters areconsidered tobetime-varying andshort- circuit failure issimulated intherotorresistance. Simulation results illustrate theeffectiveness ofthis technique.