水平变化环境下的拖线阵海底参数反演研究

Geoacoustic Inversion of Towed Line Array in Range-dependent Environment

匹配场反演(MFI)方法是快速获取海底参数的一种有效途径。常规方法采用垂直阵(VLA)可以实现海底声学参数的反演,但是其机动性较差,无法实现大面积海域海底声学参数的快速获取,而且随着声源与VLA之间距离的增大,水平变化环境将严重影响VLA的反演效果。拖线阵的优势在于能够快速机动,声源与接收阵可以同时拖动,保持较近的相对距离,从而减弱甚至忽略水平变化性对反演性能的影响。研究了利用拖线阵在水平变化环境下对海底参数进行反演的方法。反演采用低频宽带声源信号和线性匹配场处理器,并引入贝叶斯理论与蒙特卡罗方法对参数的后验概率密度(PPD)进行分析。研究结果表明:在水平变化环境下,拖线阵同样能实现海底声学参数的准确反演,并且随着声源频率数量的增多、拖线阵长度的增加,MFI结果收敛更好、可信度更高。通过计算声场传播损失并与真实海洋环境声场比较发现,反演得到的海底参数用于声场预报时,误差较小,验证了反演方法的正确性。

Matched-field inversion（MFI） is an effective method which can infer the seabed parameters quickly.Much previous works on MFI uses a vertical line array（VLA）.But,it can not be carried out in large area due to the poor mobility.Moreover,in the range-dependent environments,especially,with the increase of range between source and receiver,variation in the environment can reduce the prediction capability of the matched-field processor due to inaccuracies in the modeling.The horizontal line array（HLA） offers an advantage in terms of system mobility.The source can be towed together with the HLA,and kept at short range separations.MFI degradation due to environmental variability will be minimized or eliminated.This paper investigates the geoacoustic inversion with a towed HLA.Low frequency broad-band source signal and linear matched-field processor for geoacoustic inversion are applied,and the Bayesian and Monte Carlo methods are introduced to analyze the posterior probability density（PPD）.The PPD results indicate that the geoacoustic inversions are performed well using the towed line array,and the inversion results tend to be more convergent and credible as the number of frequencies and length of HLA increase.The transmission loss is also calculated.Compared with the true transmission loss,the result indicates that the acoustic field performance prediction by using inverted parameters has small error.