Research On Radar Target Detection Method Under Complex Clutter Background

In the simple clutter scenario,the clutter characteristics of different range cells satisfy the independent and identical distribution assumption,which can be modeled by classical clutter models,and the existing clutter suppression and target detection methods can obtain relatively ideal performance.However,under complex clutter backgrounds such as mountainous areas,cities,clouds and rain,the clutter in different areas no longer meets the IID assumption,and it is difficult to describe the clutter model exactly,which will lead to the performance loss of clutter suppression and object detection methods due to model mismatch.For example,the difference of the clutter amplitude distribution will bring about the mismatch between the target detector and the model,the difference of the Doppler spectrum will lead to the mismatch between the clutter suppression filter and the clutter Doppler spectrum.Both of the above points will seriously affect the object detection performance,and the latter has an even greater impact.At the same time,when the Doppler frequency of slow moving targets and clutter are inseparable,the target detection performance of traditional detection methods is greatly reduced.Based on the above background,this thesis focuses on the radar target detection method under the complex clutter background.The main work is as follows:1.Several clutter suppression methods and classical Constant False Alarm Rate detection methods are studied.Firstly,since several clutter suppression methods are compared,the superiority of Adaptive Moving Target Detector is explained,and the factors affecting the performance of clutter suppression are analyzed.Secondly,under the background of Rayleigh clutter,the specific expressions of false alarm probability and detection probability of several mean-level type CFAR detectors and ordered-level type CFAR detectors are derived,and the performance of the above detectors under uniform background,clutter edges,and multi-target scenes are analyzed and compared.2.Aiming at the problem of target detection under complex clutter background,refined clutter modeling and knowledge-based clutter suppression method are studied.First,with the feature of Doppler amplitude spectrum,the Gaussian mixed model is used to model the clutter of the measured data,and the clustering of the clutter background is completed.Then,based on the results of clutter clustering,clutter suppression is performed.The samples of the same cluster are used to estimate the clutter covariance matrix of the cluster,and the adaptive suppression of different types of clutter is completed.Compared with the traditional clutter suppression strategy,the knowledge-based clutter suppression method method has better clutter suppression and target detection performance.Finally,considering the timevarying characteristics of clutter,combined with the historical model parameters and the echo data of the previous frame,new model parameters are obtained by updating,and the validity of the updated model is verified by experiments.3.Aiming at the problem of slow moving target detection under ground clutter,a cognitive detection method based on clutter map in complex domain is proposed.Firstly,for the point clutter map and surface clutter map in the traditional clutter map detection method,under the background of Rayleigh clutter,the specific expression forms of its false alarm probability and detection probability are deduced,and the experiment analyzes and compares its detection performance under different parameters.Secondly,for the fixed station phased array radar,the basic principle,applicable conditions and processing flow of the proposed method are described.Finally,combined with simulation analysis and measured data processing,the performance of target detection under ground clutter background and noise background is analyzed and compared,and its performance benefits under ground clutter background are illustrated.