The retrieval of wind speed with direct detection Doppler wind lidar is performed by detecting the atmospheric back-scattering signal and computing its Doppler shift.The feasibility of applying atmospheric back-scattering signal in detection of visibility and cloud base height is analyzed.The Klett method is chosen to retrieve atmosphere visibility,and the segmentaion calculation process for cloud base height is proposed.The program based on the algorithm mentioned above is employed to process the lidar data.By comparing with the results in field test,it comes to the conclusion that the test data match the theortical analysis well.

A Fabry-Perot etalon with high spectral resolution is used as the frequency discriminator in the direct detection wind lidar.The laser frequency and the spacing of cavity drift as the temperature changes,which will cause the increasing of error.A frequency tracking solution is proposed.A backscatter fiber is used to expand the width of laser pulse.A locking filter is added to dual F-P etalon.Photo Multiplier Tube(PMT) is used to detect the intensity of light and a frequency tracking circuitry is designed.The results of simulation experiment indicated that,the spacing of cavity is able to lock to the laser by this method.

As a kind of active remote sensing tools,lidar plays an important role in atmospheric parameters measurement.According to the principle of atmospheric backscatter,the lidar for visibility measurement,namely lidar-visibility meter is designed.Firstly,the system structure is introduced in detail,and then the calibration method of return signals is illustrated.By using a new iteration algorithm which based on Fernald method to invert extinction coefficient,high precision of visibility measurement is achieved.The outfield intercomparison experiments with Belford show that the relative measurement error of lidar-visibility meter is within 20%.

Take the ultrasonic anemograph as the application background,the performance of the estimation of the time delay based on the correlation method is analyzed,and the disadvantage is pointed out.The twice correlation method,which uses the apriori information of the sending signals,is proposed as a solution.The simulation results indicate that under the same SNR,compared with the correlation method,the accuracy can be improved when using the twice correlation method.The anti-noise performance of the system has over 6 dB improvement by using the apriori information in the twice correlation method.At the end,the computation which needed in the method has been analyzed.TI TMS320C5000 series DSP processor and Altera Cyclone Ⅱ series FPGA device are taken as the hardware platform respectively.The processing time,which needed in the new method on those two kinds of platforms,are estimated.It′s concluded that the method could satisfy the real-time processing.