High-precision ranging algorithm for FMCW lidar based on improved FTRLS

In FMCW lidar ranging, noise in the signal transmission channel reduces the signal-to-noise ratio, the fence effect masks the true signal frequency, and spectral leakage leads to a decrease in frequency resolution, all of which significantly restrict the ranging accuracy. This paper proposes a refined power spectral density algorithm (RIPSD) based on the improved FTRLS (Fast Transversal recursive least squares method) to achieve high-precision ranging in FMCW lidar. This algorithm combines a redundant feedback-dynamic restart mechanism with the introduction of FTRLS to further enhance the signalto-noise ratio and incorporates a regularization factor to prevent numerical divergence. Sinc interpolation is used to refine the spectrum, solving the fence effect, and the power spectral density is applied to the radar's signal processing to improve frequency resolution and suppress interference caused by spectral leakage. Simulation experiments show that the algorithm achieves a signal-to-noise ratio improvement of more than 6 dB compared to FTRLS. To further verify the algorithm's improvement in ranging accuracy, a comparison experiment was conducted on an independently built FMCW lidar system. The results show that under the standard ranging condition of 80 meters, the average absolute error is 1.695 cm. This verifies that the algorithm has efficient noise suppression and spectral correction capabilities, providing a new highprecision ranging algorithm for FMCW lidar with broad application prospects.