Research On Digital Signal Processing Technique Of Multi-beam LFMCW Radar And Implementation

For the characteristics of no blind-zone and low transmitting power,linear frequency modulated continuous wave(LFMCW)radar is suited for near-field target detection.It is widely designed for anti-collision radar,side-scan radar and intelligent cruise control(ICC)radar in vehicle applications,while also applied in coast,dock and ship-locks monitor applications.In these applications above,the radar position is usually close to the ground or sea surface,on which position is susceptible to noise interference in target detection.Considering on the near-field target detection,the thesis proposed a kind of multi-beam LFMCW radar system.Firstly,the system parameters and design method are analysized,and then clutter module parameter estimation methods are studied.After knowing of prior knowledge of clutter,the range,velocity and angle estimation method of target in 360°azimuth field are presented under complex background.Resort to SI/PI electromagnetic simulations,a reliable high-speed digital signal processing hardware is designed.Finally,experiments are carried out to verify the rightness and robustness of system and algorithms proposed.The main contents and contributions are summarized as follows:1.A new maximum-likelihood-estimation netwton-rahpson(MLE-NR)method is proposed for clutter parameter estimation.This method is employed when the closed-form expressions of parameters in partial differential equations are not existed,it solves the selection of initial value in iteration operation for the small convergence region in traditional method.Otherwise,a modified MLE method based on damping particle swarm optimization(PSO)is proposed for K-distribution clutter parameter estimation.By setting damping boundaries,the method can reduce particle losses and increase the optimization efficiency,it is of better variance characteristics compared to traditional method.A clutter database is set up,of which data is sampled by IPIX radar and LFMCW radar designed in this thesis.Based on the database,the model recognition experiments by employing MLE-NR method and MLE-PSO method are performed and while the accurate results of goodness-of-fit to differential clutter model are given.The Monte-Carlo simulations and goodness-of-fit results demonstrate the effectiveness and robustness of this method.2.A generalized likelihood ratio test(GLRT)function of N-order AR-PSD module catastrophe is derived and a target detection method based on clutter model catastrophe theory is proposed,the method can detect the target and even can fix the appearing moment of targets in radar echo.Otherwise,it is also effective for short sample sequence(N=100).Simulations and the measured data calculation results demonstrate the effectiveness and robustness of this method.3.By knowing the prior knowledge of clutter,the ROCs of radar receiver are derived and range,velocity and azimuth angle estimation methods of target are studied.A calibration algorithm on frequency domain is employed to solve energy leakage in FFT transformation,it improves the accuracy of range/velocity estimate accuracy at the expense of little increase of computation.A 2D target estimation method is proposed,this method is suited to detect moving target under complex background or with large interferes.Simulation and experimente results demonstrate the effectiveness and robustness of this method.4.Resort to simulation software,such as allegro,ads,hfss and siwave,the SI/PI performance of high-speed signal processing hardware are analysized.Through stack-up design and resonance simulation,Z-target of power distribution network are optimized,the optimization design shows the positions,values and quantities of decoupling capacitors,which reducing the impact of rail-collapse noise,switching noise and common mode noise on radar system.Through extracting the model of vital signals,the values of matched resistors and the routing strategies are optimized,which reduce the impact of overshooting impulse,crosstalk noise and jitter noise on signal transmission.5.To investigate validity and robustness of algorithm and the hardware design,target information estimations in the 3600-scanning mode of radar are carried out.Firstly,moving target detection performance under complex background has been invested,and then multi-target detection abilities of radar system is verified.Finally,angle measurement performance of target moving inside a single beam and crossing multi beam was tested.