适用于水下高速机动目标的自适应检测方法

Adaptive Detection Method for Underwater High-speed Maneuvering Targets

近年来,发展主动声呐目标检测技术逐渐成为水声信号领域的研究热点。该技术在实际应用中,浅海探测环境的复杂多变导致辅助数据获取困难,同时目标高速运动时引起的距离徙动现象会导致水下目标检测性能下降。针对这类问题,提出了一种适用于水下高速机动目标的自适应检测方法。首先,构建多元假设检验模型来准确描述高速机动目标的回波分布情况。其次,使用模型阶数选择方法来估计目标回波位置。然后,利用干扰协方差矩阵的斜对称特性有效减少辅助数据的需求量,以降低检测方法对辅助数据的依赖性。最后,基于广义似然比检验准则实现未知参数的最大似然估计以及检测统计量的推导。仿真结果表明,该方法具有恒虚警特性,同时在辅助数据受限的情况下,相较于其他同类检测方法有7dB以上的目标检测性能优势,并且能够更准确地估计目标回波分布情况,有效改善了距离徙动现象导致的检测性能下降问题。

In recent years,the development of active sonar target detection technology has gradually become a research hotspot in the field of underwater acoustic signals.In the practical application of this technology,it is difficult to obtain auxiliary data due to the complex and changeable shallow sea detection environment,and the detection performance of underwater high-speed maneuvering targets is degraded because of the phenomenon of Range Cell Migration(RCM).Aiming at such problems,an adaptive detection method for underwater high-speed maneuvering targets is proposed.Firstly,a multiple hypothesis testing model is constructed to accurately describe the distribution of high-speed maneuvering targets echoes.Secondly,the Model Order Selection(MOS)method is used to estimate the target echo position.Then the persymmetric structure of the disturbance covariance matrix is used to effectively reduce the amount of auxiliary data required,in this way the dependence of the detection method on auxiliary data is reduced.Finally,the Maximum Likelihood Estimate(MLE)of unknown parameters and the derivation of detection statistics are realized based on the Generalized Likelihood Ratio Test(GLRT)criterion.The simulation results show that the proposed method has the characteristics of Constant False Alarm Rate(CFAR).At the same time,when the auxiliary data is limited,the proposed method has a target detection performance advantage of more than 7dB compared with its counterparts,and can more accurately estimate the target echo distribution,the problem of detection performance degradation caused by distance migration is effectively improved.