Data Processing And Short-range Millimeter Wave LFM Detection Target Recognition

The Millimeter-wave linear frequency modulated continuous wave (LFMCW) radar has a wide range of applications in the short-range detection field with its advantages of high range resolution, low-emission power, no range blind of area, et al. Based on a pratical project, the paper studies on range-velocity coupling problem, high-resolution range profile and target classification and recognition algorithm, and has developed a target processor to achieve real-time signal processing.Firstly, based on the working principle and the beat-signal characteristics of the LFMCW radar, the range-velocity coupling problem has been studied; the MTD technology, velocity estimation technology based on the plus and minus slope of frequency modulation and MTD-Frequency domain pairing method are focused; the velocity estimation technology is adopted to eliminate the range-velocity coupling; Simulation results are given.Secondly, methods to improve ranging accuracy of LFMCW radar are studied based on the imaging principle of high-resolution range; such as the padding zero method, Chirp-Z transform method and Zoom-FFT method based on polyphase filter structure; After discussion of these algorithms’ computation and implementation, the padding zero method is chosed as the back-end processor’s real-time implementation algorithm; Against an impact of the target motion having on high-resolution range profile, the velocity compensation method is adopted based on the plus and minus slope of frequency modulation; Simulation results show that the method is effective.Thirdly, the problems of gesture sensitivity, shift sensitivity and magnitude sensitivity of high-resolution range profile are discussed, and the corresponding pretreatments are given. Against the influence on orientation sensitivity problem bringing to high-resolution range profile, the preprocessing method of average range profile is given; Simulation results show that the method is effective; The method of target feature extraction based on scattering centers is focused, and the results of linear discriminant function used for target classification are analyzed; Simulation results have proved the effectiveness of the target recognition algorithm.Finally, the hardware and software implementation of the target processor is studied. According to the signal processing requirements, FPGA is chosed as the processor chip, and it is divided into data acquisition section and signal processing section. The detail of each module circuit design is focused. On the aspect of software implementation, Nand Flash controller of data acquisition and the target recognition algorithm implention of signal processing are focused, each sub-module is simulated. With the help of PC software, the debugging of data acquisition and signal processing is completed, which shows that the target processor has high performance and it meets the requirements of short-range detection.