Adaptive Control for Periodical Time-Varying Systems with Unknown Frequencies

Abstract

This paper presents a self-tuning technique designed to handle the periodical fast time-varying systems. The parameters of the system vary rapidly with unknown fequencies and amplitudes. We only know the form of variation is sinusoid. A frequency detector, which consists of a combined multistage estimator, a filter, a peak detector and a monitoring program, is used to estimate the frequencies of time-varying parameters. The estimated frequencies in turns are used in the estimator with time-varying model to obtain the fine estimates of the time-varying parameters which will be used to compute the parameters of adaptive controller. The result of this combination produces a satisfactory control performance unattainable otherwise. It also shows the feasibility of tailoring the existing adaptive control algorithm to fit the special type systems.