基于同态滤波技术的水下目标运动参数估计

针对测量平台空间有限条件下的近场目标运动分析问题，提出了一种基于浅海射线声学多途结构的单水听器水下目标运动分析(TMA)方法。通过构建匀速直线运动目标的三维多途时延模型，推导了以运动参数为变量的直达声和海面一次反射声时延差的非线性函数表达式；根据典型水声信道特征，推导了信道的倒谱表达式，进而利用同态滤波技术解卷积能力估计直达声和海面一次反射声时延差；结合Levenberg-Marquardt方法求解非线性方程的最优解；最后提出二次逼近校正方法，降低了实际水文条件对参数估计的影响。仿真结果证明了方法的正确性，参数估计精度满足实际应用需求。

Estimation of Underwater Target Motion Parameters Based on Homomorphic Filtering

The eanalysis of target motion in near field is difficult when the space of measurement platform is limited. A novel underwater target motion analysis (TMA) method is proposed based on the multipath structure of the ray acoustics in shallow water, A three-dimensional multipath time delay model of target in uniform speed rectilinear movement is established. The nonlinear function expression of time delay between direct sound and surface reflected sound is derived by taking the movement parameter as a variable. Based on the characteristics of typical underwater acoustic channel, the expression of channel in cepstrum domain is achieved. Then the time delay between direct sound and surface reflected sound is estimated by the homomorphic filtering technique with the ability of deconvolution.The optimal solution of the nonlinear equation is solved by combining with Levenberg-Marquardt method. The quadratic approximation correction method is proposed to decrease the influence of actual hydrological condition on parameter estimation. The simulation results show that the proposed method is practicable, and the estimation precision meets the demand of practical application.