Research And Implementation Of Vehicle Detection Radar Signal Processing System Based On TMS320C6678

With the large-scale construction of expressways,intelligent traffic monitoring systems have more and more applications.The highway vehicle detection radar system can monitor traffic conditions in real time,record vehicle information,illegally capture and alarm,which can effectively improve the level of highway operation and management.Based on a vehicle detection radar as a project background,this paper studies the radar signal processing algorithm for the distance movement and speed ambiguity problems faced by the radar when detecting vehicle targets,and designs and implements a real-time highway vehicle target based on highperformance multi-core DSP TMS320C6678 Software solution for detection.In this paper,the research is carried out in the order from theoretical algorithm verification to interface design and debugging,to the realization of the multi-core solution,and finally to the system joint debugging.Combined with the project indicators,the following parts are mainly discussed:1)Research on signal processing algorithms.A highway vehicle target echo model is established.Aiming at the large bandwidth of this model of radar transmission signal,there will be echo distance walking and speed ambiguity problems when detecting moving targets.A Keystone transform method based on velocity ambiguity number search is adopted to obtain When the target information is unknown,the effect of correcting the echo distance walking and solving the speed blur.In order to improve the performance of radar target detection under clutter and noise interference,the averaging unit constant false alarm detection algorithm is simulated,and on this basis,the averaging unit selects the constant false alarm detection algorithm,which can effectively improve the clutter edge effect,filter noise and Clutter and other signals;compared traditional target aggregation and connected-domain target aggregation algorithms,connected-domain algorithm has the advantages of saving algorithm running time and solving the problem of target splitting in traditional target aggregation;single pulse and differential amplitude angle measurement methods are analyzed,and The simulation results are given to determine the target orientation information.2)TMS320C6678 communication interface scheme design and debugging.In response to the need for high-speed data transmission between DSP,FPGA,and upper computer,a radar system communication interface scheme was designed,and software design methods for GPIO,SRIO and Gigabit Ethernet interfaces were given,and GPIO,SRIO and Gigabit Ethernet were completed.Debug the network interface and determine that the transmission rate of SRIO in4 X mode is 11.34 Gbps,and the network port rate is 602 Mbps,which meets the communication requirements.3)Realization of multi-core solution.Because single-core implementation cannot process radar echo signals continuously at high speed,a software implementation of multi-core,masterslave ping-pong processing based on SYS/BIOS real-time operating system is proposed,and EDMA is used to move data,and Message Q is used to achieve inter-core communication and Solve the problem of Cache consistency.Through this solution,real-time processing steps such as constant false alarm,target detection,speed blur resolution,and angle measurement are completed,and target information reporting is completed every other CPI to meet the needs of real-time detection of highway vehicles.4)System software debugging.The echo is simulated by the radar simulator,and the modules of the DSP inter-core communication interface,EDMA,constant false alarm,target aggregation,and angle measurement are debugged.The consistency of the DSP processing results and the simulated target information is compared,and the field measurement results are given.,Verified the correctness,real-time and stability of the whole system.The vehicle detection radar signal processing algorithm and DSP signal processing system studied in the thesis meet the project requirements,have the reliability and real-time performance of continuous detection,can complete the vehicle detection,and lay the foundation for the successful development of the project.