Research On Positioning Technology Based On Radio Interferometry

Owing to its passive measurement,high angle measurement accuracy,long operating distance and so on,Interferometry technology is suitable for the near space non-cooperative low-velocity target position determination which flying in above the ground from 20 to 100 kilometers,and becomes a hotspot research.Theoretical analysis and simulation experiment were implemented,aiming at the problems of radio interferometric positioning of non-cooperative targets in the near space,including positioning principle,positioning error analysis,target position optimization,etc.The main contents of this thesis are as follows.Firstly,the interferometry technology and its advantages on positioning non-cooperative target in the near space are introduced.According to the principle of interferometry positioning technology,the mathematical model is established and solution algorithm is analyzed.Different positioning evaluation criteria are compared.Secondly,the positioning equation of non-cooperative target interferometry is established,and the calculation process of geometric accuracy factor is derived.The influence of measurement time error,baseline length and station location error on the geometric accuracy factor is simulated and analyzed with different receiving station configurations.The Gauss Newton method is used to calculate the target position.The results show that the root mean square error of the target in the short baseline system of 1000 meters is close to 45m with the time difference error 5ns and the station location error 0.02m,while in the small array baseline system of 10 meters is close to 18m with the time difference error 0.02ns and the station location error 0.01m.The shorter the baseline length,the greater the time difference error;the higher the target height,the lower the positioning accuracy of the target;the long baseline system is less affected by the site error,while the short baseline system is more affected by the site error.Then,a Levenberg-Marquarelet algorithm solution model and Intelligent algorithm solution model are established.The capable of Calculating the target position and the advantages of low requirement of initial value and high robustness of LM algorithm are verified by numerical simulation.Finally,in order to solve the problem of slow convergence speed and many iterations of basic genetic algorithm in solving the target position,an adaptive genetic algorithm strategy and a hybrid strategy combining genetic algorithm and Gauss Newton algorithm are proposed.Simulation results show that the improved the iteration times of adaptive genetic algorithm is reduced from 190 to 80 times and hybrid algorithm have about 30 times faster convergence speed from 11.56 to 0.374 second when 1000 solving the non-cooperative target position.