Research On Inversion Method For Evaporation Duct Based On GPS Satellite Signal

The atmospheric duct is an abnormal atmospheric refraction phenomenon in the ocean.The propagation trajectory of radio wave is different from the standard atmospheric environment,and there is an abnormal propagation phenomenon,which seriously affects radar ranging,velocity measurement and so on.The propagation of radio wave in the atmospheric duct layer is with small energy attenuation and can realize over-the-horizen propagation.It can effectively solve the problem that limited range of radar wave in propagation.The problem is caused by the protruding of the earth surface,and the exisitence of atmospheric duct can make radar detect the targets outside the horizon effectively.However there may be some electromagnetic wave leakage at the top of the duct to make the target not be captured.In other words,it will also lead to the blind area of radar detection,which will affect the propagation of radio wave and lead to the failure of radar detection target location.Therefore,the in-depth study of atmospheric duct can further guide the improvement the problems in other radio electronic systems,such as radar detection,positioning to capture targets,communication and so on.According to the above introduction it is also explained that studying on atmospheric duct is trategic and decisive significance in military high-tech electronic warfare.The appearance of evaporation duct is universal and has become the most concerned type of atmospheric duct.The research of over-the-horizon detection in coastal,ship radar and other practical applications is mainly focused on evaporation duct environment.With the development of remote sensing and satellite communication technology,it is possible to use GPS satellite signal to retrieve evaporation duct profile.Compared with traditional detection methods and radar sea clutter detection,it has the characteristics of strong real-time performance and wide prediction area.In this paper,the basic characteristics of atmospheric duct are introduced which conclude the formation mechanism,generation conditions,classification and causes of atmospheric duct.Meanwhile,the frequency and wavelength conditions of atmospheric duct propagation are analyzed.The direct detection methods of evaporation duct and the prediction methods of characteristic parameters of evaporation duct based on marine hydrometeorological parameters are briefly described,analyzed and evaluated.A new technique for expanding the measurement range of evaporation duct is proposed,and the new technology is briefly described and analyzed.Then,on the basis of analyzing the principle of GPS signal propagation,the parabolic equation of GPS signal propagation in evaporation duct is derived in detail,and its numerical algorithm is given detailly.Meanwhile,with emphasis on the initial field,upper and lower boundary conditions and split-step Fourier algorithm in the parabolic equation of evaporation duct based on GPS signal inversion.After that,the propagation loss of GPS signal in evaporation duct is simulated and analyzed.The inversion method of atmospheric duct based on GPS signal is also presented.Secondly,neural network BP and genetic algorithm GA,are introduced and applied into the inversion process of evaporation duct.Based on the analysis of the practical application of these two methods,an improved GA-BP neural network algorithm for evaporation duct inversion is proposed.The inversion principle,steps and flow chart are given and the formula for calculating the propagation power distribution is derived.Meanwhile,the nonlinear relationship between the propagation power of GPS signal and the characteristic parameter of evaporation duct,that is,evaporation duct height,is established.Finally,the improved GA-BP neural network algorithm is simulated,and the evaporation duct height and evaporation duct profile under the BP neural network algorithm and the GA optimized algorithm are simulated,and the comparison and analysis are carried out.The results show that the two algorithms are feasible for the inversion of evaporation duct using GPS signal,but the evaporation duct profiles inversed by GA-BP neural network algorithm fit well with the theoretical profiles,and the error is small.After further noise inversion,the GA-BP neural network algorithm is closer to the actual section,which shows that the evaporation duct profile has good robustness and generalization ability.At the same time,it is shown that the optimal initial weight and threshold of BP neural network can be obtained effectively by combining GA with BP neural network,which improves the accuracy of predicting evaporation duct based on GPS signal inversion.