Generalized Adaptive Notch Filter With a Self-Optimization Capability

The paper presents a self-optimizing version of a generalized adaptive notch filter (GANF). Generalized adaptive notch filters are used for identification/tracking of quasi-periodically varying dynamic systems and can be considered an extension, to the system case, of classical adaptive notch filters. The tracking properties of a GANF algorithm depend on two adaptation gains, which should be chosen so as to match the degree of nonstationarity of the identified system. First, an analytical study of a tracking performance of a GANF algorithm is presented. Then, based on the obtained theoretical results, a self-optimizing GANF algorithm is proposed, capable of automatic tuning of its adaptation gains