Research On Target Detection Algorithm Of Ground Reconnaissance Radar And DSP Realization

Ground surveillance radar,which played a great role in the war years is an important means of obtaining personnel and vehicle information on the battlefield.In the new era,the widely using robots and UAV are characterized by slow moving speed,strong concealment and small radar reflection area.It is difficult for traditional ground surveillance radars to effectively monitor such small slow targets.In this paper,with the demand of the ground surveillance radar detecting of slow small target,we carry out research on clutter suppression and slow target detection algorithm,and complete the development of radar target detection and system control module upon DSP system.The main work of this paper is as follows:(1)Clutter suppression technology research.Based on the study of the principle of clutter generation and its statistical characteristics,the amplitude distribution model and power distribution model of a variety of radar clutters are analyzed.The zero-memory nonlinear transformation method is used to simulate the receiving clutter of ground-based radar.Based on this,clutter suppression techniques such as MTI,MTD and clutter map has being discussed and the advantages and disadvantages of each method has being analyzed.(2)Radar target detection technology.First of all,we discuss a VI-CFAR detector.Then analyze the application of PHD filter based on particle filter in target detection.Finally,we propose a CFAR-PF-PHD detection method based on CFAR detection and the method uses a double-threshold VI-CFAR detector to pre-process and classify the data.Then,the classified data is reduced dimensioned and the suspect target is identified and tracked by using PF-PHD.The simulation results show that the method has high precision,low computational complexity and good performance of slow target detection.(3)Target detection algorithm in multi-core DSP implementation.Aiming at the functional requirements of the radar target detection module,the DSP system architecture of the target detection and system control module is completed.The multi-core parallel computing is used to realize the target detection algorithm and optimize the DSP program to improve the system's execution efficiency and meet the complex algorithm Real-time requirements.The system has been successfully applied to a new type of foreign trade portable ground surveillance radar.