Research And Hardware Implementation Of Highway Radar Signal Processing Algorithm

With the large-scale construction of highways,the number of cars is constantly increasing,which has brought tremendous pressure to traffic vehicle monitoring.Due to the high speed and mobility of highway vehicles,the traditional radar detection method will have the problem of distance migration in long-term accumulation,it is difficult to achieve accurate target detection,and it cannot meet the detection accuracy requirements of traffic monitoring systems.Therefore,it is necessary to study the radar signal processing method applied to highway targets,which has important civilian and engineering practical significance.In this paper,the corresponding signal processing algorithm is given based on a highway radar project,and based on the hardware architecture of FPGA + DSP,a scheme for highway radar signal processing based on TMS320C6678 DSP is designed and implemented.This paper aims at the target detection in the highway scene.Firstly,through mathematical derivation,a highway target echo model is established.According to the theoretical formula and MATLAB simulation,the problem of distance migration under the traditional radar signal processing are analyzed,and the distance migration of the echo signal is proportional to the target velocity,signal bandwidth,accumulation time and other parameters.Secondly,this paper adopts a Keystone transform algorithm based on parallel search of velocity ambiguity numbers.Under the condition that the target velocity is unknown,the distance migration of the highway target can be effectively corrected,and the velocity ambiguity problem can be compensated accordingly,and the signals can be accumulated for a long time based on this method.Compared with the traditional method,the accumulation gain of the echo signal-to-noise ratio is improved,and the detection performance of the highway target by the radar is improved.Through constant false alarm,target condensation,sum-difference angle measurement,the distance,speed,and azimuth of each vehicle on the highway are detected.Then,under the conditions of the entire engineering system,the data transmission interfaces such as the SRIO and network port of the DSP 6678 were debugged,and the correctness and reliability of data transmission between the DSP and the FPGA and between the DSP and the host computer were verified.Next,the entire signal processing flow is implemented based on a single core,and the resources and time occupied by each algorithm module are analyzed,and the signal processing scheme based on 7 cores is designed and implemented.Among them,Core 1receives the processed sum difference beam data processed by the FPGA through the SRIO interface.Core 0 is responsible for network port communication,receives radar signal parameters from the host computer,and sends the target detection results after the entire signal processing to the subsequent data processing module.Core 2～core 6 is used for constant false alarm,target condensation and angle measurement of sum and difference ratio under different velocity ambiguity number of multi-beam signal.Finally,multi-core communication and multi-core optimization were designed,and algorithm resources and time were further optimized,verified the consistency of the signal processing results with MATLAB simulation.The entire signal processing flow meets the system’s real-time requirements.This paper implements a radar signal processing algorithm applied to the detection of highway targets,which effectively solves the problems of distance migration and speed ambiguity in traditional radar detection methods.The design and implementation of a hardware scheme that satisfies real-time performance and high processing performance is given,and the detection of highway targets is completed.