Convergence rates of an adaptive control algorithm with application to the speed control of a DC machine

Abstract

A method of obtaining the convergence rates of the minimal-controller synthesis (MCS) strategy, as applied to a first-order SISO (single-input-single-output) plant, is described. The MCS algorithm is a significant development of model reference adaptive control, and is based on the hyperstability theory of Popov. Bounds are provided on the speed of convergence of the closed-loop error dynamics. It is shown that no knowledge of the plant dynamics is required, apart from an estimate of the low-frequency gain. An estimate of the plant output settling-time is derived. The problem of MCS speed control as applied to a DC machine is investigated from an error-convergence viewpoint.<>