Node Self-Localization Technology Research Of Distributed Through-the-wall Radar

The distributed through-wall radar system uses multiple node radars with different perspectives to detect hidden targets inside buildings or building layout structures collaboratively,which overcomes the shortcomings of traditional single-station radar information acquisition by acquiring information from multiple perspectives and completing effective integration.Therefore it has important application value and social significance in the fields of anti-terrorism reconnaissance,disaster rescue,environmental perception and so on.Compared with the traditional single-station through-wall radar,the radar nodes of the distributed through-wall radar system are usually deployed depend on the building structure,and multiple radar nodes detect collaboratively.Therefore,the research on the high-precision positioning of radar nodes in the distributed through-wall radar system is of great significance for subsequent radar data fusion.Aiming at the positioning problem of distributed through-wall radar nodes,this paper researches the positioning method based on ultra-wideband,binocular vision,and radar image registration,and realizes the distribution by three different methods,the specific content is as follows:1.In the positioning method based on the ultra-wideband system,based on the working principle of the ultra-wideband system and the working characteristics of the distributed through-wall radar system,a reasonable base station deployment method and positioning method are designed to ensure that the positioning system has a higher positioning accuracy.For the actual ultra-wideband system,the system ranging calibration and positioning performance analysis are carried out.Finally,a positioning system based on MATLAB is developed,and indoor and outdoor positioning experiments are carried out.The results show that outdoor positioning has high positioning accuracy.Indoor positioning is subject to multipath,the positioning effect is not as good as outdoor;2.In the positioning method based on the binocular vision system,according to the positioning principle of the binocular system,the influence of various parameters of the binocular system on the positioning accuracy is studied and a reasonable system structure is designed.The method of image segmentation,extracting the centroid of the object is designed for radar node detection.and Z.Z.You calibration method is applied to calibrate the internal and external parameters of the camera,and finally the effectiveness of the radar node positioning method based on binocular vision is verified through experiments;3.In the positioning method based on radar image registration,each radar node can obtain imaging information based on its own coordinate system.The imaging information contains the position of targets in the scene.According to the position information of targets,the position of the radar node can be acquired.Therefore,the imaging results of each radar node are preprocessed first,and then the graph model is constructed.The same target under different views is correlated according to the graph matching algorithm,and the radar position is finally solved by associated target position.The proposed method is verified by electromagnetic simulation and measured data,and the results show the proposed method has higher positioning accuracy;The above-mentioned distributed through-wall radar node positioning methods have been verified by actual measurement data,and have been proved to have high indoor positioning accuracy,which can be applied to practical application.