A THEORETICAL AND EXPERIMENTAL COMPARISON OF THE ITERATIVE EQUIVALENT SOURCE METHOD AND THE GENERALIZED INVERSE BEAMFORMING

Many acoustic source mapping methods exist to perform noise source localization and quantification and appear to be powerful tools for acoustic diagnosis in industrial applications. Two classes of methods have known many developments in the last few decades: deconvolution algorithms combined with beamforming (CLEAN, DAMAS, etc) and inverse methods such as the Equivalent Source Method (ESM) and the Generalized Inverse Beamforming (GIB). In this paper, a special attention will be paid to the use of inverse methods in complex acoustic environments. Recently, Suzuki has demonstrated the applicability of the GIB to the study of aerodynamic sound sources [25], highlighting comparable performances to the existing deconvolution techniques. On the other hand, an iterative version of the ESM has been proposed in the context of acoustic imaging in closed spaces, at INSA Lyon [22]. This paper provides a theoretical and experimental comparison between two inverse methods: the iterative ESM and the GIB using various benchmark problems and aeroacoustic experimental data. The experimental set-up consists of a steel rod placed in the potential core of a rectangular jet inside an open-jet anechoic wind tunnel. It will be shown that both methods are based on similar mathematical formulations although they were developed for different application fields. Reconstruction performances of the algorithms in terms of localization and quantification will be discussed as well as their computational efficiency.