Characteristic Moore-Greitzer model parameter identification for a one stage axial compressor system

Axial compressor systems are predisposed to instability near their optimum operating point. Instabilities include surge or stall, leading to severe consequences to the operational health and integrity of compressor system. The Moore-Greitzer (MG) model has been commonly recognized as a standard when characterizing the dynamics within an axial compressor and is advantageous for the development of a controller. Such a controller promises to increase the efficiency of compressor systems; yet, controller design has been barred by an inability to extract the MG parameters defining the behavior of real-life compressors. Hence, control has not been based on the MG model. Determining these system parameters experimentally is impractical due the limited range of operation compressors can withstand without sustaining damage. In this paper a proof-of-concept gray-box identification method is proposed to extract the characteristic parameters of a MG model from experimental data. This technique utilizes a genetic algorithm based optimization. In this study, simulated data from a MG model and measured data from a one stage compressor system is utilized to extract key parameters of the MG model. Establishing an indirect method to determine the parameters for the MG model extends its relevance from theoretical use to concrete application and opens the door for the direct control of axial compressors.