Coherent Integration Method Based on Analytical Numerical Model and Local Interpolation for Distance and Velocity Detection in FMCW LiDAR

To address the issue that range migration correction is inapplicable in FMCW LiDAR caused by Doppler shifts much larger than the laser repetition frequency, an improved method based on analytical numerical models and local interpolation is proposed. Firstly, theoretical analysis shows that the range migration correction only depends on the time-bandwidth product multiple of the FMCW LiDAR system. Therefore, the ambiguity number formula is modified by establishing a relationship between the time-bandwidth product multiple and the modified factor. Secondly, the local interpolation method enables precise alignment between the maximum coherent integration value and the velocity estimation of the modified ambiguity number formula, thereby achieving range migration correction. Then, coherent integration and one-dimensional constant false alarm rate are applied to the corrected data, significantly reducing the computational load compared to the traditional methods. Finally, it is experimentally demonstrated that, compared with the traditional method, the improved method achieves more than 5 dB improvement in beat frequency signal-to-noise ratio for moving targets with velocities exceeding 100 m/s, and reduces the computational load by more than 90% when the integration period exceeds 500.