| With the continuous development of global navigation satellite system,satellites play a vital role in military defense and civil fields.However,the electromagnetic environment is increasingly severe,satellites are particularly vulnerable to unintentional or anthropogenic interference,which poses a threat to the safety of satellites.Faced with the interference of these radiation sources on the satellite,the research of satellite interference source positioning technology has the key practical significance and necessity.This paper mainly uses the principle of arrival time difference to locate the interference source,studies a variety of time difference parameter estimation algorithms,and simulates the multi-satellite time difference of arrival model.The main research contents of this paper are as follow :Analyzed several common models that use different parameters and different numbers of satellites to locate interference sources.Based on the feasibility,this paper selects the multisatellite time difference positioning model.After deriving the positioning performance evaluation index,the result shows that the positioning performance is not only related to the random error of the time difference estimation,but also by the system error such as the relative position error of neighboring stars.Improving the time difference estimation algorithm is a key step in improving the positioning accuracy.For the problem of parameter estimation,some algorithms are studied:time difference estimation based on cross-correlation function,time difference estimation based on cross power spectrum,and time difference estimation based on adaptive filtering.However,the accuracy of time difference estimation still needs to be improved under the conditions of correlated noise and low SNR.This paper proposes a variable step size adaptive filtering algorithm based on the fourth-order cumulant,which can not only effectively suppress the influence of correlated Gaussian noise on the positioning accuracy,but also improve the convergence speed and the estimation accuracy in a weak environment.Using the prior knowledge that the ground interference source is located on the surface of the earth,three-star time difference positioning model based on perfect sphere and WGS-84 ellipsoid is simulated.The iterative algorithm is studied,and the simulation results show that the Newton iteration method is better than the spherical iteration method in terms of running time and positioning accuracy.And through Hessian matrix,iterative step factor,the Newton iterative method is modified.In addition,in order to eliminate the system error to improve the positioning accuracy,four reference stations are set on the ground to further calibrate the position of the interference source.Under the same system error,the positioning accuracy of the calibration algorithm has been significantly improved.For non-terrestrial interference sources,we can increase the number of satellites to complete the location of the interference source.In addition to the iterative and analytical methods,this paper introduces the flower pollination algorithm into the time difference positioning model,and improves it from the perspectives of initialization,conversion probability,and search strategy.The simulation results show that the improved flower pollination algorithm has improved convergence performance,and the localization performance is better than the iterative method and the analytical method. |
