Abstract: Estimating the target position of low-frequency sound sources in a shallow sea environment is difficult due to the high cost of hydrophone placement and the complexity of the propagation model. We propose a compressed recurrent neural network (C-RNN) model that compresses the signal received by a vector hydrophone into a dynamic sound intensity signal and compresses the target position of the sound source into a GeoHash code. Two types of data are used to carry out prior training on the recurrent neural network, and the trained network is subsequently used to estimate the target position of the sound source. Compared with traditional mathematical models, the C-RNN model functions independently under the complex sound field environment and terrain conditions, and allows for real-time positioning of the sound source under low-parameter operating conditions. Experimental results show that the average error of the model is 56 m for estimating the target position of a low-frequency sound source in a shallow sea environment.

一种基于数据驱动的浅海低频声源目标位置估计方法

[1]Agarwal A, Kumar A, Agrawal M, 2015. Iterative adaptive approach to DOA estimation with acoustic vector sensors. OCEANS, p.1-8.

[2]Bharathi BMR, Mohanty AR, 2018. Underwater sound source localization by EMD-based maximum likelihood method. Acoust Aust, 46(2):193-203.

[3]Chai T, Draxler RR, 2014. Root mean square error (RMSE) or mean absolute error (MAE)?—Arguments against avoiding RMSE in the literature. Geosci Model Dev, 7(3):1247-1250.

[4]Chao A, Shen TJ, 2003. Nonparametric estimation of Shannon’s index of diversity when there are unseen species in sample. Environ Ecol Stat, 10(4):429-443.

[5]D’Spain GL, Luby JC, Wilson GR, et al., 2006. Vector sensors and vector sensor line arrays: comments on optimal array gain and detection. J Acoust Soc Am, 120(1):171-185.

[6]Demuth HB, Beale MH, Jess OD, et al., 2014. Neural Network Design. Martin Hagan.

[7]Dushaw B, 2014. Acoustic Tomography, Ocean. In: Njoku EG (Ed.), Encyclopedia of Remote Sensing. Springer, New York, USA, p.4-11.

[8]Felisberto P, Santos P, Jesus SM, 2010. Tracking source azimuth using a single vector sensor. Int Conf on Sensor Technologies and Applications, p.416-421.

[9]Fox A, Eichelberger C, Hughes J, et al., 2013. Spatio-temporal indexing in non-relational distributed databases. Int Conf on Big Data, p.291-299.

[10]Gu JX, Wang ZH, Kuen J, et al., 2018. Recent advances in convolutional neural networks. Patt Recogn, 77:354-377.

[11]Hildebrand JA, 2009. Anthropogenic and natural sources of ambient noise in the ocean. Mar Ecol Progr Ser, 395:5-20.

[12]Ian G, Yoshua B, Aaron C, 2016. Deep Learning. The MIT Press, USA.

[13]Li GY, Kawan B, Wang H, et al., 2017. Neural-network-based modelling and analysis for time series prediction of ship motion. Ship Technol Res, 64(1):30-39.

[14]Li TC, Su JY, Liu W, et al., 2017. Approximate Gaussian conjugacy: parametric recursive filtering under nonlinearity, multimodality, uncertainty, and constraint, and beyond. Front Inform Technol Electron Eng, 18(12):1913-1939.

[15]Moussalli R, Srivatsa M, Asaad S, 2015. Fast and flexible conversion of Geohash codes to and from latitude/ longitude coordinates. Proc IEEE 23^rd Annual Int Symp on Field-Programmable Custom Computing Machines, p.179-186.

[16]Porter MB, 2011. The BELLHOP Manual and User’s Guide: Preliminary Draft. Technology Report, USA.

[17]Praczyk T, 2015. Using evolutionary neural networks to predict spatial orientation of a ship. Neurocomputing, 166:229-243.

[18]Prieto A, Prieto B, Ortigosa EM, et al., 2016. Neural networks: an overview of early research, current frameworks and new challenges. Neurocomputing, 214:242-268.

[19]van Gerven MAJ, Bohte SM, 2017. Editorial: artificial neural networks as models of neural information processing. Front Comput Neurosci, 11:114.

[20]Wittekind D, Schuster M, 2016. Propeller cavitation noise and background noise in the sea. Ocean Eng, 120:116-121.

[21]Zhao AB, Ma L, Ma XF, et al., 2017. An improved azimuth angle estimation method with a single acoustic vector sensor based on an active sonar detection system. Sensors, 17(2):412.

[22]Zhou JB, Zhang MH, Piao SC, et al., 2019. Low frequency ambient noise modeling and comparison with field measurements in the South China Sea. Appl Acoust, 148:34-39.