Simulation Design And Data Processing For The Satellite Hybrid Doppler Wind Lidar

Combining the advantages of coherent and direct detection for Doppler wind lidar(DWL),the space-based hybrid DWL reduces the limitation of height measurement range by direct or coherent detection alone,promises to reduce technology demands and investment in “very large” individual DWLs,becomes the optimum solution for global tropospheric wind field detection.In this paper,we study the wind field detection at the altitude of 0?30km by the space-based hybrid DWL.Firstly,simulation design for the spaced-based hybrid DWL system and its echo signal are carried out according to technical specifications of the spaced-based detection for wind field inversion.Secondly,the data processing algorithms are studied for the coherent Doppler wind lidar(CDWL)echo signal.Then the three-dimensional wind field inversion is performed on the simulated echo signal of the spaced-based hybrid CDL.Finally,a demonstration GUI interface for 3D wind-field inversion is compiled for each module of the spaceborne hybrid wind lidar system.1000 sets of simulation signals are processed using the hybrid inversion technique,and the spatial and temporal distribution map of the inversion results are performed.The main creative work done in the thesis is as follows:1.Aiming at the uneven of the signal power spectrum baseline of coherent detection in the far-field for the spaceborne hybrid systems,an auto-adaptive background subtraction(AABS)algorithm is proposed.Then AABS algorithm,Fast Fourier Transform(FFT)algorithm and adaptive iterative weighted penalty least squares(air PLS)algorithm are used to process the simulation data and experimental data.The results show that the AABS algorithm has better noise suppression effect and outperforms the FFT algorithm and air PLS algorithm in the furthest detectable range,which has improved 40% and 16.7% respectively.2.Due to the low SNR in the far-field detection by the CDWL,weighted subspace fitting(WSF)algorithm based on signal subspace decomposition are proposed.The proposed WSF algorithm and Eigen Vector(EV)algorithm are verified by simulation signal and experimental data.The processing results show that the proposed algorithm can reduce the statistical uncertainty of the estimated wind velocity and improves the detection range approximately up to 26.6% and 14.2% respectively when compared to the FFT algorithm and the EV algorithm.3.In order to solve the problem that the maximum likelihood algorithm needs to predict the parameters for coherent echo signal processing,the maximum likelihood algorithm based on covariance matrix is proposed.Two algorithms are used to process the simulated data and the experimental data respectively.The processing results show that the maximum likelihood algorithm based on the covariance matrix does not need to predict the spectral width and signal-to-noise ratio parameters,avoids the error caused by the estimated parameters,and has a higher reliability of unbiased estimation and a larger detected range,which is improved by20% compared with the maximum likelihood algorithm with a priori knowledge of spectral width.4.According to the requirements of simulated echo signals for the spaceborne hybrid lidar wind field detection,the 3D wind field inversion technique are studied and the interactive simulation software for integrated echo signal simulation,wind speed error analysis and wind speed inversion are compiled.Among them,the wind field inversion is carried out by the coherent heterodyne technique at the altitude of 0 ? 6 km,and direct double-edge technology at the altitude of 3 ? 30 km.The hybrid inversion technique are used to process 1000 sets of simulation signals,and the spatial and temporal distribution map of the inversion results are analyzed.The analysis results show that the designed spaceborne lidar system meets the technical requirements that the wind speed error is less than 3 m/s and the wind direction error is less than 10°.