Parameter Identification of Compressor Dynamics During Closed-Loop Operation

Abstract

A low-speed axial research compressor has been fitted with movable inlet guide vanes to allow for feedback stabilization of rotating stall. A model exists whose structure captures the input-output behavior, and stabilization of rotating stall is possible using this model. Quantitative identification of the parameters in the rotating stall model requires the ability to identify MIMO dynamics, which may be unstable, during closed loop operation. The ‘instrumental variable’ technique is presented as the basic approach to this problem. The necessary extensions to the basic technique are discussed, and the resulting algorithm is applied. Experimental results are presented which verify that the methodology yields useful estimates.