Assessment of sensorless Kalman filter-based dual-loop control strategies for grid-connected VSIs with LCL filters

H. M. El-Deeb, A. Massoud, A. S. Abdel-Khalik, S. Ahmed

Research output: Chapter in Book/Report/Conference proceedingConference contribution

2 Scopus citations

Abstract

This paper illustrates the assessment between the multiloop control strategies that aim to damp the resonance hazard of LCL-filters and expanding the available stability margins. The establishment of the multiloop control is sustained in a sensorless manner with the reconstruction of the state variables using the Kalman observer, thus enhancing the cost measures and reliability of the distributed generation (DG) interface. Compared to the conventional multiloop techniques, which is mainly based on the feedback of either the filter capacitor current or voltage, the proposed multiloop method is able to exhibit: 1) theoretical stability through the whole frequency spectrum; 2) higher disturbance rejection originating from the grid-side interference; 3) higher damping of the dominant poles with much lower control effort than the conventional methods; 4) straightforward tuning of the grid current controller as the LCL-filter appears as an apparent standard second order model; and 5) low sensitivity against parameter deviations compared to previous literatures. Furthermore, the utilized control scheme ensures a high quality of injected grid currents under unbalanced and distorted grid voltage conditions. Experimental results validated the proposed control concept through a 3.5 kW setup.
Original languageEnglish (US)
Title of host publication2015 IEEE Energy Conversion Congress and Exposition, ECCE 2015
PublisherInstitute of Electrical and Electronics Engineers Inc.
ISBN (Print)9781467371506
DOIs
StatePublished - Oct 27 2015
Externally publishedYes

Fingerprint

Dive into the research topics of 'Assessment of sensorless Kalman filter-based dual-loop control strategies for grid-connected VSIs with LCL filters'. Together they form a unique fingerprint.

Cite this