Fibonacci array-based focused acoustic camera for estimating multiple moving sound sources

Abstract

Multiple moving sound sources’ aliasing problem occurs in many fields. The number and relative position of these sources provide critical information for further signal processing. However, estimating the number and relative position under aliasing of multiple moving sound sources is challenging. In this study, the Fibonacci sequence was applied to design a novel two-dimensional Fibonacci array, which has lower maximum sidelobe level than other typical array structures in a certain frequency scope. Conventional beamforming was utilized as an acoustic camera for stationary sound source localization by two-dimensional spatial scanning. Unlike the acoustic camera, a spatiotemporal map (STM) method was developed to estimate the number and relative position of multiple moving sound sources through spatial scanning and temporal focusing. STM is called focused acoustic camera because of the spatiotemporal information extraction of dynamic sources. A deconvolution approach called clean STM (CSTM) was created to estimate the spatiotemporal parameters for further signal processing and to overcome the influence of array pattern on STM. CSTM was combined with two-dimensional time-varying spatial filtering rearrangement and harmonic product spectrum-based sparse filtering enhancement. Consequently, a systematic technique solution was developed to be potentially applied in wayside acoustic bearing fault diagnosis systems. Simulation and experimental cases verified the effectiveness of the proposed method and the systematic technique solution.