Research On Target Detection Algorithms In Non-uniform Sea Clutter Background

As an important branch of radar target detection,target detection technology in non-uniform sea clutter has attracted more and more scholars,attention.With the improvement of modem radar by researchers,the development of high-resolution radar system has been promoted,which makes sea clutter non-Gaussian and non.uniform in time and distance dimensions.At this time,the likelihood ratio detector based on the characteristics of uniform sea clutter can not detect the target accurately.Based on this,this paper studies the target detection algorithm under the background of non-uniform sea clutter,so as to improve the detection performance of the likelihood ratio detector.The main contents of this paper are as follows:(1)The normalized sample covariance matrix estimation algorithm for clutter power is studied.Aiming at the problem that the deteetion performance of the detector is degraded due to the existence of abnormal elements of high-power target-like signals in non-uniform sea clutter.Based on the power mean normalized covariance matrix estimation algorithm,an improved power mean normalized sample covariance matrix(Men-NSCM)estimation algorithm is proposed in this paper.Firstly,by means of clutter power mean processing,the anomalous elements of target-like signals have smaller weights,so as to reduce their influence on the estimation of sample covariance matrix.Then,a new likelihood ratio detector is constructed by combining the improved covariance matrix estimation algorithm with the traditional adaptive filter detector.Finally,the new detector constructed by the improved covariance estimation algorithm has better detection performance in clutter environment with high-power target-like signals by comparing and analyzing the measured data with the detector constructed by the traditional covariance estimation algorithm.(2)A target detection algorithm based on inverse composite Gauss distribution model is studied.In order to improve the detection performance of Likelihood Ratio Detector in non-uniform sea clutter environment.Firstly,through the analysis of the statistical characteristics of sea clutter,it is verified that the inverse composite Gauss distribution model ean better describe the statistieal characteristics of non-uniform sea clutter data.Then,based on this distribution model,the structure of inverse composite Gauss distribution likelihood ratio detector with maximum posteriori estimation is deduced without increasing computational complexity.Finally,by comparing the likelihood ratio detector with the traditional inverse composite Gauss distribution detector and the measured data,it is proved that the detector has better detection performance under the same conditions.(3)Inverse Compound Gauss Distribution Detector(ICGD)is used to study the effect of clutter power.In order to improve the detection performance of the likelihood ratio detector based on inverse composite Gauss distribution model in clutter environment with high power target-like signals.Firstly,the normalized covariance matrix estimation algorithm based on the influence of clutter power is introduced into the detector,and four new likelihood ratio detectors are constructed.Then,the new constructed detector and the traditional adaptive inverse composite Gauss likelihood ratio detector are compared and analyzed under the same conditions,which verifies that the new constructed inverse composite Gauss likelihood ratio detection based on the influence of clutter power mean has stable detection performance in non-uniform and non-Gauss clutter environment.