基于声场模信号特征和多项式拟合的声速剖面反演技术研究

Sound Speed Profile Inversion Based on Mode Signal and Polynomial Fitting

针对未知环境下利用经验正交函数表示声速剖面的局限性,提出了一种基于多项式拟合和模信号相结合的声速剖面反演方法。介绍了模信号的概念和深海分布特性,通过理论分析和仿真分析,证明低阶多项式拟合可用于深海声速剖面的表征,并提出采用5阶多项式拟合以实现对2000 m以浅的声速剖面较精确的表征。在反演处理流程中,利用不同参数的多项式拟合计算获得拷贝模信号,利用脉冲声信号获得数据模信号,并通过遗传算法进行参数寻优。以典型Munk剖面为例仿真分析基于垂直阵、单水听器、非海面海底反射(SRBR)信号的反演性能,单水听器的声速剖面反演与垂直阵性能接近,并提出基于非SRBR模信号反演可以提升在海底参数未知条件下反演的技术思路。利用4000 m海深的单水听器接收的脉冲信号(距离180 km,接收与发射深度均为100 m)进行方法性能验证,结果表明,共轭深度以下(100~2000 m)的声速剖面反演误差小于0.2 m/s.

A sound speed profile inversion method based on mode signal and polynomial fitting is proposed for the limitation of using the empirical orthogonal function to represent the sound speed profile in unknown environment.A concept of mode signal and the distribution characteristics of deep sea are given.It is proven through theoretical analysis and simulation that the low-order polynomial fitting can be used for presenting the sound speed profile in seabed.The five-order polynomial fitting can be used for accurately presenting the sound speed profile in 2000 m-depth sea water.And in the inversion process,the copy mode signal vector is obtained by polynomial fitting of different parameters,and the mode signal is obtained from the test data by using the pulse acoustic signal,of which the parameters are optimized by genetic algorithm.The inversion performances of vertical array,single hydrophone and non-SRBR signal are simulated and analyzed by taking typical Munk profile as an example.It is concluded that the higher the order of polynomial is,the better the fitting effect of sound speed profile is.When a certain error is allowed,the lower order can be used for fitting the sound speed profile.The simulated results show that the performance of sound speed profile inversion using a single hydrophone is close to that of the vertical array,and the inversion performance under the condition of unknown seabed parameters can be improved by using the non-SRBR mode signal inversion.And finally,the performance of the proposed method is verified by using the pulse signal received by the single hydrophone in 4000 m-depth sea water(distance:180 km,and receiving and transmitting depth:100 m).The results show that the inversion error of sound speed profile is less than 0.2 m/s.