Research On Parameter Identification Of PMSM Servo System

Due to the wide application of rare earth materials servo system, the permanentmagnet synchronous motor (PMSM) is becoming mainstream.When a PMSM servosystem is at work, because of two issues of the heat and the magnetic saturationphenomenon, motor parameters will gradually deviate from the original values withdifferent fluctuations. It is really a hard task for the latest high-precision servo drivecontrollers; therefore, PMSM servo controllers with function of parameter identificationis leading all the products. The main research work and conclusions of the paper are asfollows:At first, the PMSM vector control strategies and its analytical model were analyzedand the structural relationship PMSM servo control system is well-studied;Next, the off-line parameter identification methods for their specific embodimentsoffline parameter identification has been deeply analyzed for PMSM stator resistance,inductance andAC-DC axis of the permanent magnet flux.Then, according to the principle of least squares, and the advanced algorithm, anew kind of online parameter identification method has been researched and simulationunder the Matlab/Simulink software environment is well-conducted. Different results ofparameters identification in various situations are compared. Given the shortcomingsFFRLS, PMSM motor model with the variable inertia is re-analyzed according toMRAS algorithm. It is proved that the FFRLS has great performance, but MRAIalgorithm is less dependent on the initial terms. As a result, the convergence of MRAI inthe initial uncertainty still in accepted error range, especially in the situation of timevariance.Finally, to combine the time-varying parameter identification results withdouble-loop controller auto-tuning strategy, taking into account the variability PMSMservo system is a strong high-end systems, the current loop is derived variable speedloop works real-time multi-parameter tuning PI tuning The method and the simulation,the speed waveform observation verifies the reliability of the method.