Research On Radar Target Adaptive Detection Algorithm Under Sea Clutter

In view of the non-uniform and non-Gaussian environments under sea elutter,when there are abnormal units with large power,the traditional covariance matrix estimation has poor estimation accuracy due to strong clutter interferenee.In order to overcome this limitation,the existing The adaptive detector is improved,and the effectiveness of the proposed method is verified by simulation and measured data.The specific content of this paper is as follows:(1)Target detection is based on the statistical properties of sea clutter.Firstly,from the statistical point of view,this paper analyzes the amplitude distribution models of several different conditions and applicable ranges by using IPIX measured radar data,and verified that the K distribution under the composite Gaussian model can best reflect the statistical characteristics of real sea clutter.The SIRP method generates a composite K random sequence,which is the basis for subsequent simulation experiments.(2)Firstly,adaptive detection algorithms and covariance matrix estimation methods are reviewed.Then,for the ease of sea elutter environment unevenness,the accurate estimation of the covariance matrix is affected by the power anomaly in the abnormality unit.The method of multiplying the clutter covariance matrix estimation of the unit to be detected by the power median of the elutter reference unit to obtain an improved covariance estimation method,that is,the normalized sampling covariance matrix of clutter information of joint unit to be detected based on power median(PM-MNSCM.It is finally verified by simulation and measured data.(3)In order to further improve the performance of radar adaptive target detection algorithm in non-uniform clutter environment based on the composite Gaussian model,the proposed PM-MNSCM estimation method is applied to the adaptive detection algorithm(α-AMF),which depends on the shape parameter.Therefore,the a-AMF detector based on SCM-PMMNSCM combination estimation method is designed.The simulation data and measured data are used to compare the original a-AMF and the improved α-AMF.The performance analysis results can be obtained in non-uniform non-stationary Under the composite Gaussian model,the performance of the proposed detection method outperforms the comparison detection algorithm.