A river velocity inversion method based on a multi-peak precise matching algorithm for acoustic tomography

Abstract

This paper addresses signal matching challenges encountered in River Acoustic Tomography (RAT) [1] technology applications within short-distance rivers. To overcome limitations associated with significant errors in traditional maximum peak methods within such environments, a novel multi-peak accurate matching approach is proposed. This method integrates a greedy strategy [2] with the Hungarian algorithm [3], resulting in enhanced accuracy. Experimental findings demonstrate that velocity inversion errors are reduced to below 10% when compared to conventional manual flow measurement techniques, representing a substantial improvement. Furthermore, a comprehensive technical framework for RAT technology, spanning from signal processing to parameter inversion, is presented. This framework offers a viable solution for unmanned, real-time intelligent hydrological monitoring using RAT technology. This study promotes the application of acoustic tomography [4] from ocean to river, and provides a new technical paradigm for hydraulic engineering and ecological monitoring.