Online-Oriented Smoothing Iterative Learning Control for FMCW LiDAR Nonlinearity Compensation

Nonlinear compensation technology is the key to frequency-modulated continuous-wave (FMCW) light detection and ranging (LiDAR) systems. Among compensation algorithms, the iterative learning control (ILC) pre-distortion algorithm has gained much attention for its high accuracy and simplicity. However, the conventional ILC algorithm can only be used for offline calibration because it cannot work for long-term stable operation. To address this problem, this paper proposes a smoothing ILC to compensate for the nonlinearity of FMCW LiDAR. The experimental results show that the method substantially improves the convergence stability (99.5%) and the convergence accuracy (96.5%). Meanwhile, the method optimally reduces the frequency sweep relative residual nonlinearity to 0.0004% and residual nonlinearity to $2.1 \times {{10}^{ - 10}}$. To the best of our knowledge, this method achieves the highest nonlinear compensation effects. Furthermore, the results of the robustness and ranging experiments show that the method is not only highly robust but also improves the ranging accuracy by 48%.