Research On Signal Processing Algorithm Based On Vehicle Mounted Millimeter Wave Radar

With the rapid development of the world economy and high-tech,driverless technology has gradually become the new direction of vehicle design in the future.Unmanned driving technology needs to rely on high-precision on-board sensors.Among them,the vehicle millimeter wave radar is an important part of the current vehicle sensors.The core of the vehicle millimeter wave radar system is its signal processing module.The signal processing algorithm based on digital signal processing technology is of great importance and far-reaching significance to the research and development of vehicle-borne millimeter-wave(MMW)radar.This paper discusses the development of MMW radar on vehicle,summarizes the development and realization of signal processing algorithm of MMW radar on vehicle,and points out the problems in traditional signal processing algorithm of MMW radar on vehicle.The overall structure and basic working principle of the vehicle millimeter wave radar system are analyzed.The working principles of the millimeter wave radar with different operating frequencies and various radar modulation waveforms are studied.Through analysis and comparison,finally,24 GHz triangular wave linear frequency modulation continuous wave radar is taken as the research object.Aiming at the problems of low resolution,false alarms,and failure to detect correct targets in traditional signal processing algorithms,a complete signal processing scheme is proposed.The Fast Fourier Transform(FFT)is used to convert the time domain signal to the frequency domain.At the same time,windowing in the time domain is performed to prevent spectrum leakage.A CZT(linear Z transform)algorithm is proposed to improve the resolution.On the basis of this,the resolution of the spectrum can be greatly improved to better identify the target.,an improved CFAR detection algorithm is proposed and feedback is introduced Mechanism for the problems of false alarm and missing alarm in the traditional CFAR(constant false alarm)algorithm,which can solve this problem well.The variable-period LFMCW + CW(linear frequency modulation continuous wave + constantfrequency continuous wave)modulation waveform is chose to process for multi-target pairing algorithm,the distance can be solved well and the problem of speed coupling can reduces the existence of false targets and correctly identifies target objects.TMS320F28335 DSP processor is chosen as the main chip of the signal processing module,and the implementation of the related algorithms in the signal processing module is described in detail.The ADC module,DMA module,low-pass filter module and serial port module in signal processing module are analyzed.Finally,a joint test of MATLAB and DSP was carried out.The speed and distance results obtained after single-target and multi-target simulation tests met certain error requirements,thereby verifying the feasibility of the signal processing algorithm proposed in this paper.According to the simulation results,the next research direction is proposed.In this paper,the research on signal processing algorithm of vehicle-borne MMW radar is helpful to improve the accuracy of radar identifying target information,and promote the performance of vehicle-borne MMW radar system,and also have certain guiding significance for the MMW radar-based automatic driving research.