Application Of Joint Time-frequency Analysis In Coherent Doppler Wind Lidar

Coherent Doppler lidar,with high accuracy,high spatial-temporal resolution and large-scale coverage,has been widely applied to the detection of wind shear,aircraft vortex,atmosphere turbulence,wind power generation and so on.The key issue for lidar signal processing is to extract weak Doppler frequency shift in the weak backscatter signal.Compared with time-domain and frequency-domain methods,joint time-frequency analysis describes the full view of signals in time-frequency domain and reveals the time-varying feature by the joint density function of the time and frequency.Therefore,it is easy to detect low-level signal submerged in noise by time-frequency analysis and multiple accumulation.Also,joint time-frequency analysis can maintain the continuity of wind speed in far field and enhance the temporal resolution and range resolution.Furthermore it can track the wind details better and improve signal-to-noise.The mainly work of this thesis are as follows:1.The current progress of the coherent Doppler lidar and joint time-frequency analysis are reviewed,and the appilcations of time-frequency analysis in lidar are followed,namely the future development tendency of time-frequency analysis for lidar.The principle of atmospheric wind measurement based on Doppler effect is analysed,and basic elements of coherent lidar system are introduced.The atmospheric slices model for coherent Doppler lidar is established,and common algorithms such as pulse pair algorithm,maximum likelihood estimator,and periodogram maximum estimator are descibed in detail.The performance:parameters of coherent lidar are resumed,the basic theory of time-frequency methods is discussed.2.Based on the atmospheric slices model,the echo signal of coherent Doppler lidar is simulated.The simulated signal of coherent Doppler lidar is processed by linear time-frequency representation,generalized bilinear time-frequency representation,reassigned time-frequency representation,the adaptive optimal-kernel time-frequency representation.Particulary time-frequency feature and the capability of various methods are compared.3.The spectrogram,adaptive optimal-kernel time-frequency representation and periodogram maximum estimator are applied to the field experiment data derived from a 1.5?m coherent Doppler lidar in Hefei,Anhui Province in March,2017,and the retrieved wind velocity results are compared.The velocity ranges from-3.5m/s to Om/s,and wind shear around 1.1 kilometers is formed due to wind field Canyon effect.Experimental results show that adaptive optimal-kernel time-frequency representation can track the wind details better and maintain the continuity of wind speed in far field when the range resolution is 1.2m and temporal resolution is setted to 0.2s.