| Clouds,which are small water droplets or ice crystals formed by the evaporation of water,are common meteorological phenomenon in nature.60%-70%of the earth's surface is covered by clouds.They have significant influence on global radiation balance,atmospheric energy budget and hydrological cycle.At present,common cloud detection devices include laser ceilometer,radiosonde,etc.,but due to their limited detection abilities,it is hard to obtain more information about cloud properties.In recent years,with the advantages of stability,good penetrability and high resolution,millimeter-wave cloud radar has become the focus of industrial field's research.This thesis starts from the selection of millimeter-wave cloud radar detection modes,and studies the cloud echo processing algorithm by setting appropriate parameters.The main research contents and conclusions are as follows:(1)The first part firstly outlines the system composition of the millimeter wave cloud radar used in this thesis,including the hardware system and the software system.Then,according to the characteristics and structure of the cloud signal,four detection modes are designed to meet the detection requirements in different scenarios.Mode A is used for low-altitude observation and can meet the 300m blind zone requirement.Mode B has the best signal-to-noise ratio gain in the middle and low areas.Mode C is used for mid-high altitude stratus and cirrus detection.And mode D has the farthest detection distance and the highest signal-to-noise ratio gain.Many experimental results indicate that the combined operation of the four modes can cover almost all the cloud change process,ensuring the comprehensiveness of the detection task.Finally,the radar receiver is calibrated to ensure its stability and reliability.(2)The second part introduces the algorithm design of cloud radar system,including echo signal processing algorithm and power spectrum processing algorithm.The former includes pulse compression,coherent integration,FFT and incoherent integration,while the latter mainly includes noise level calculation,cloud signal recognition,spectral moment calculation and quality control.In terms of the calculation of precipitation cloud noise level,based on previous studies,this thesis takes the power over the longest distance as the reference noise level.Then,from high to low,according to the atmospheric attenuation values corresponding to different rainfall intensities,the noise level of the precipitation cloud in low altitude is calculated.This method obtains more accurate results through experiments.(3)The third part studies the relationship between accumulation techniques and the signal-to-noise ratio gain.Accumulation technology is an important method used to improve the signal-to-noise ratio in meteorological radar.This thesis analyses the echo phase to obtain the relationship between coherent integration numbers and Doppler frequency,thereby the most suitable scheme can be selected in different scenarios.Incoherent integration is mainly used for frequency spectrum data superposition.Thus,by analyzing the degree of linear correlation between spectra and spectra,this thesis proposes the correlation coefficient method,which means more accumulation while highly correlated and less accumulation while poorly correlated.The two schemes mentioned above have achieved good results in the tests. |
PDF Full Text Request