Inversion Technology Of Regional Sound Velocity Profile Based On Inverted Echo Sounder

With the steady implementation of the marine power and smart ocean strategy,underwater acoustic observation and monitoring are gradually developing towards the direction of threedimensional and precise data acquisition.Sound velocity profile has obvious temporal variability characteristics,which have an important impact on sound signal propagation,and then affect the working performance of sonar systems such as underwater acoustic communication,positioning,and mapping.Finding an inversion method that can accurately and quickly obtain the temporal variability characteristics of sound velocity profile has become a current research focus.The aim of this article is to enhance the underwater acoustic detection capability in an environmentally coupled system by improving the sonar system’s perception of ocean environmental parameters.Based on the characteristics of the inverted echo sounder(IES)to achieve large-scale and long-term dynamic monitoring of the marine environment,a series of studies were conducted on the problem of using multiple IESs to jointly invert the regional sound velocity profile.By inverting the sound velocity profile through sound propagation time and combining with existing data,research on IES-based regional sound velocity profile inversion technology was conducted using data processing methods,data assimilation,Gravest Empirical Mode,spatial interpolation,and empirical orthogonal function methods.The research content of this paper is mainly divided into three parts.The first part is the principle and data preprocessing technology of IES.The second part is the sound velocity profile inversion technology based on IES.The third part is the reconstruction technology of regional three-dimensional sound velocity field.In the first part,in view of the problem that there are missing,abnormal,and noisy data in the original data of IES,missing value interpolation,abnormal value detection and removal,and smoothing filtering are used to process the original data of IES,and high-quality data that meet the usage standards are obtained.Then,the long-short-term memory network is used to predict the sound propagation time series data,and the predicted results combined with the sound velocity inversion technology can predict the sound velocity profile at future moments.In the second part,a sound velocity profile inversion method based on IES is studied.Firstly,the Gravest Empirical Mode(GEM)method is used to project the historical hydrological data in the research area onto the space of sound propagation time functions,establish the empirical relationship between historical hydrological data and sound propagation time at each standard depth layer,and conduct an overall study on historical hydrological data.Then,the ensemble Kalman filter is used to assimilate multiple source hydrological data to obtain more accurate hydrological data,and finally,the assimilated GEM field is constructed to improve the accuracy of sound velocity inversion.The third part uses multiple IES deployed at different locations and assimilated historical hydrological data to invert the sound velocity GEM field at multiple discrete points in space.Various spatial interpolation methods are used to reconstruct the regional 3D sound velocity profile field.The inverted sound velocity profile data at multiple discrete points are decomposed into several basic functions using empirical orthogonal functions.The 3D sound velocity field in the regional area is constructed by spatially interpolating the coefficients of the basic functions,which reduces the complexity of interpolation operations and achieves the rapid establishment of the 3D sound velocity profile field.Finally,the error of various spatial interpolation methods for interpolating the original data and EOF time coefficients to reconstruct the 3D sound velocity field is compared using the leave-one-out cross-validation method.The results show that compared with other spatial interpolation methods,the spatiotemporal kriging method has the best interpolation effect in both the reconstruction of the3 D sound velocity field based on the original data and the one based on the 3D sound velocity field represented by empirical orthogonal functions.