Improved apFFT Algorithm Processing and Hardware Implementation for FMCW LiDAR
To address issues in the backend signal processing of Frequency Modulated Continuous Wave(FMCW)LiDAR,such as the picket-fence effect,spectral leakage,and processing speed bottlenecks of traditional FFT-based frequency estimation algorithms,an improved all-phase FFT(apFFT)time shift phase difference algorithm with high frequency estimation accuracy and processing speed is proposed.This method effectively combines the phase characteristics of the Digital Down Conversion(DDC)output signal with the initial phase invariance of the apFFT algorithm.By leveraging power characteristics to reduce the FFT calculation points,the frequency estimation process is accelerated.Additionally,the power-based spectral offset compensation parameters are applied to the apFFT phase difference method to further enhance frequency estimation accuracy.Simulations and FPGA validations demonstrate that,with 1 024 sampling points,the proposed algorithm reduces the root-mean-square error(RMSE)by 85%compared to traditional FFT methods,with an average frequency estimation error of less than 100 Hz.A complete FMCW ranging and velocity measurement system was imple-mented for field experiments.Statistical analysis of 10 000 measurements shows that the standard deviation of ranging results is less than 8 mm,with a minimum of 0.58 mm.The standard deviation of velocity measurements is less than 0.04 m/s,with a minimum of 0.008 7 m/s.The frequency estimation rate reaches up to 25 kHz.
FMCW LiDARDistance and speed measurementHigh speed and high precisionReal-time processingall-phase time shift phase difference
万方数据
维普
