基于AR模型的自适应浮动门限检测方法

An Adaptive Threshold Detection Based on AR Model

混响是水下目标回波信号检测的主要背景干扰;基于AR模型的预白化匹配滤波检测方法,往往由于混响AR模型定阶困难,无法对混响准确建模;这将会导致预白化效果不佳,检测性能下降;针对这一问题,在预白化匹配滤波的基础上,提出了一种基于AR模型预白化匹配滤波的自适应浮动门限检测方法;给出了浮动门限经验公式,并对实测数据进行处理;处理结果表明,所提出的基于AR模型预白化匹配滤波的自适应浮动门限检测方法能使检测门限很好地跟随信号的变化趋势,提高检测概率,保持恒虚警检测。

Reverberation is the dominant background interference in the short-distance under water target echo detection.As it is difficult to estimate the accurate auto regressive（AR） model parameters,the detective method based on the pre-whiten matched filter using auto regressive（AR） model is not very effective.It leads to the capability of the detection decreases rapidly.Focused on the problem,an adaptive threshold adjusting method based on the AR model is proposed in this paper.And then,the experience formula of adaptive threshold is given out and tested on real reverberation data.The result shows that the method proposed in this paper can follow the trendline and improve the detection capability and maintain the Constant False Alarm Rate（CFAR） detection.