On the Enhancement of Generalized Integrator-based Adaptive Filter Dynamic Tuning Range

Hafiz Ahmed, Mohamed Benbouzid

Research output: Contribution to journalArticle

Abstract

Many phase locked-loop (PLL) or frequency locked-loop (FLL) use quadrature signals (directly or indirectly) as the input variables. Generalized integrator (GI) is a popular quadrature signal generator (QSG) available in the literature. GI is also widely used in various industrial applications. In addition to being a QSG, GI also works as kind of band-pass filter. However, due to structural limitation, the dynamic tuning range is limited for the standard GI. The limitation arises from using only one gain in the direct-phase estimation dynamics while quadrature phase estimation dynamics doesn’t use direct feedback of the filter estimation error. Some attempts have already been made to overcome this limitation by adding direct feedback of the filter estimation error to quadrature phase dynamics as well. However, we have demonstrated in this paper that this kind of implementation has some frequency domain limitations. In this paper, we propose a novel GI type adaptive filter using coordinate transformation. The resulting structure maintains the same kind of filtering property of the standard GI at the transformed coordinates level while at the same time enhances the dynamic tuning range of standard GI. Details of the proposed technique, stability analysis and discussion on gain tuning are provided in this paper. Finally, comparative experimental results are provided with respect to GI-FLL to show the dynamic performance improvement. Experimental results demonstrate the suitability of the proposed technique.
Original languageEnglish
Pages (from-to)(In-Press)
Number of pages9
JournalIEEE Transactions on Instrumentation and Measurement
Volume(In-Press)
Early online date20 Mar 2020
DOIs
Publication statusE-pub ahead of print - 20 Mar 2020

Fingerprint Dive into the research topics of 'On the Enhancement of Generalized Integrator-based Adaptive Filter Dynamic Tuning Range'. Together they form a unique fingerprint.

  • Cite this