Nonlinear adaptive control of permanent magnet step motors

Abstract

The problem of accurate positioning of permanent magnet step motors with sinusoidal flux distribution is considered. On the basis of a nonlinear model containing six unknown parameters, a nonlinear adaptive control is designed that guarantees asymptotic tracking of a desired angle reference signal, feeding back the whole state measurements (position, speed and currents). The algorithm may be used for preliminary on-line identification of those parameters that do not vary during operation: persistency of excitation conditions are given that guarantee identification with exponential convergence. When the only unknown parameters are load torque and stator resistance, which vary during operations, persistency of excitation conditions are satisfied when stator currents are not zero: in this case the adaptive control provides on-line exponential parameter estimation and exponential tracking of desired position signals. Simulations show the performance of the adaptive algorithm and a comparison with PID control.