To solve the problems of complex equipment in traditional Doppler wind LiDAR, high energy requirements of single-transmitter dual-receiver laser detection, and significant blind zones, this study proposes a direct path-averaged crosswind detection system based on the laser scintillation cross-correlation method. The system adopts an integrated single-ended dual-channel transceiver structure using a 1550 nm laser. By constructing multi-layer turbulent phase screen models for simulation and validating with a dual-channel crosswind detection system, the proposed method achieves path-averaged crosswind detection while reducing peak power requirements and enhancing robustness under complex meteorological conditions.
In the theoretical study, Mikhail Belenkii's analysis was fully referenced. In a round-trip path, the laser beam propagates through atmospheric turbulence twice: From the transmitter to the target and then back to the receiver. When the system parameters meet the requirements, the intensity fluctuations of the reflected beam on the round-trip path are mainly determined by scintillation along the return path. Based on the above theory, simulations were conducted using the power spectrum inversion method to generate atmospheric turbulent phase screens. A multi-layer vacuum-phase screen alternating transmission model was constructed based on the van-Karman spectrum. The intensity fluctuation characteristics of laser propagation were calculated using a wave optics model. The theoretical feasibility of the scintillation cross-correlation method was verified by comparing the simulation results with the correlation calculation results. In the experimental section, a dual-channel coherent laser detection system was developed using an innovative coaxial dual-transmitter dual-receiver structure. Atmospheric scintillation signals were collected using a 1550 nm laser and high-speed photodetectors. During the data processing phase, signal analysis was performed using a combination of spectral energy extraction algorithms and the zero-mean normalized cross-correlation (ZNCC) algorithm.
Simulation results revealed a linear relationship between the cross-correlation peak offset of the echo energy in the round-trip path and the crosswind speed, with a relative error of 3.2%, demonstrating strong consistency and verifying the accuracy of the theoretical model (Fig. 5). In the experimental phase, well-designed and reliable comparative tests were conducted. A comparison between the experimental data and reference values from a Doppler wind LiDAR (Fig. 10a) showed consistent trends in wind speed fluctuations, with a Pearson correlation coefficient of 0.865. Linear regression analysis (Fig. 10b) indicated a determination coefficient R2 = 0.688 between the experimental data and reference values, with a fitting slope approaching 1 and a root mean square deviation (RMSD) of 0.755 m/s. Residual analysis (Fig. 11) validated the normal distribution of errors (K-S test, p > 0.05), and the symmetric distribution of residuals indicated that the system exhibited stable statistical characteristics.
Through numerical simulations and experimental verification, this study confirms the effectiveness of the single-ended scintillation cross-correlation method for path-averaged crosswind detection. The innovative coaxial dual-transmitter dual-receiver structure not only reduces the peak power requirements but also enhances system reliability. The high correlation (r = 0.865) between experimental results and standard radar data demonstrates that the system can provide reliable data support for meteorological monitoring and engineering applications, laying a foundation for further optimization of algorithm adaptability under different meteorological conditions.
DownLoad: