Minimal controller synthesis algorithms with output feedback and their generalization

Abstract

This paper proposes several improvements in the minimal controller synthesis algorithms, which were developed for a class of nonlinear systems with uncertainties. The major proposition is that only the output feedback is enough to control some nonlinear systems without an observer while the existing algorithms require the complete state feedback. Next, the extended version and the parameter identification technique of the minimal controller synthesis algorithm are combined in a single method estimating some more parameters for some known forms of nonlinearities. This is applicable with or without the improvement of removing the need for using the complete state feedback. It is proven that the scaling factors in the algorithms can be variable and different for each component. The algorithms are applicable to some noncanonical forms of systems. All the propositions and some of the existing algorithms are generalized in a single form with some options. A speed-sensorless DC motor control simulation is also included with quite exaggerated noisy conditions and the results verify that the proposed modifications make the minimal controller synthesis algorithms much more efficient model reference adaptive control algorithms for certain types of systems. The proposed method, which does not use the speed measurement, yields fewer ripples than the existing method, which uses the speed measurement. In addition, the proposed method accurately estimates the armature resistance and inductance while the existing method fails to do this.