Adaptive Observer of Synchronous Motor State Variables with Unknown Parameters and Measurement Noise

The problem of synthesizing an adaptive observer of the state variables of a permanent magnet synchronous motor under conditions of parametric uncertainty and noise in stator current measurements is considered. A parameterization of the motor model is proposed, which makes it possible to obtain a regression model with unknown motor parameters and constant offsets in stator current measurements. The Kreisselmeier scheme is applied and devices for identifying unknown parameters are constructed to ensure convergence in a finite time. The generated estimates are used by an adaptive observer to reconstruct the angular position of the engine. A nonlinear dynamic model of a permanent magnet synchronous motor is analyzed. It is assumed that the only measurable signals are stator currents and voltages, while the currents contain unknown constant offsets. The only known parameter of the motor is the inductance of the stator windings. Using dynamic filtering, a linear regression model is derived for several unknown parameters, including the resistance of the stator windings and constant measurement noise. A dynamic extension of the regression model using the Kreisselmeier scheme is caried out. The conditions are formulated under which the constructed devices for identifying unknown parameters ensure guaranteed convergence of estimation errors in a finite time. Using the estimates obtained, an adaptive observer of the total magnetic flux, angular position, and rotational velocity of the rotor is synthesized. The results of the computer simulation demonstrate the proposed observer effectiveness for a typical operation scenario of motor to which an unknown load moment is applied.