Experimental verification of adaptive and robust trajectory tracking controllers for switched reluctance rotors

Integrator backstepping design techniques have previously been used to develop voltage level control algorithms for various classes of electric machines. The resulting control algorithms yield rigorous stability results that are attractive for motor drive applications for several reasons: (i) they can compensate for typically neglected electrical dynamics associated with the motor drive system, (2) they can compensate for a variety of uncertainties throughout the entire drive system (i.e., both modeled and unmodeled electromechanical dynamics), and (3) they can be designed to operate with only partial state measurements (i.e., sensorless control). Despite this promising outlook, few backstepping based algorithms have been experimentally verified, primarily due to their recent theoretical emergence. This paper examines two backstepping based algorithms for switched reluctance (SR) machines which use full-state measurements (i.e., motor position, velocity, and per phase winding currents), to yield rigorous theoretical stability results on the rotor trajectory tracking error despite uncertainty in the SR motor drive system. The controllers incorporate a smooth commutation strategy for the shared specification of the individual SR phase currents. The proposed controllers are experimentally verified using a digital signal processor based data acquisition and control system, and the performance results compared to a backstepping based, exact model knowledge control algorithm