Equivalent source method based near field acoustic holography using multipath orthogonal matching pursuit

Abstract

Recently several equivalent source method (ESM) based near field acoustic holography (NAH) methods have been proposed based on compressive sensing, such as one norm convex optimization (L1CVX) and iteratively reweighted least squares (IRLS). The compressive sensing methods have the advantage of reconstructing the fields well beyond the spatial frequency limit. In the paper, multipath orthogonal matching pursuit (MPOMP), developed from matching pursuit (MP) and orthogonal matching pursuit (OMP), is proposed to predict the source location better. The study examines the performance of MP, OMP, MPOMP, L1CVX, and IRLS for multiple monopoles across a range of frequencies. MPOMP, L1CVX, and IRLS methods are studied and compared numerically for a vibrating shell and three loudspeaker acoustic field configurations. The performance of MPOMP is also examined experimentally for three loudspeaker configurations and compared with L1CVX and IRLS. MPOMP achieved better source localization and pressure field prediction than L1CVX and IRLS. Also, in terms of computational time, MPOMP is much more efficient than L1CVX and IRLS.