Research On Direction Finding Techniques Of Missile-borne Radar

Because of the limitations of the special application environment and hardware scale,some high performance but computationally expensive algorithms can not be applied to the missile-borne radar.However,the missile and target relative ultra-high-speed flight requires the Direction-Finding(DF)system to have accurately real-time estimation performance.Focusing on the background of the anti-aircraft missile and airborne target relatively ultra-high-speed movement,this thesis researched the DF techniques of high-accuracy and highly real-time based on missile-borne radar.The contents are divided into four parts shown as follows.Firstly,this thesis introduced the guidance model of semi-active seeker,analyzed the narrow-band noise and the continuous-wave signals that often appearing in the semi-active seeker of air defense missiles.Then,studied the two-dimensional(2D)interferometer system that is suitable for real-time direction-finding,meanwhile introduced the principle of direction finding in 2D interferometer.By analyzing the error of direction finding in 2D interferometer,the factors affecting the accuracy of estimation are given.Secondly,the closed-form direction-finding algorithm based on uniform circular array,which is suitable for radar seeker,is studied and deduced.After that,this thesis analyzed the measuring principle and measurement error of phase difference based on Fast Fourier Transform(FFT).In order to reduce the computation complexity of measuring phase difference using FFT,studied the effect of grouping FFT on the phase difference estimation of both accuracy and computation.The conclusion is that grouping FFT can get the same DF accuracy as non-grouping,but it can greatly reduce the amount of computation.On the basis of the above analysis,also introduced the method of solving the phase ambiguity using parallel baselines.And under the jamming of Gaussian white noise and channel phase uncertainties,analyzed the conditions required for the correct unwrapping phase ambiguity of parallel baselines in detail.Inspired by the neural network and machine learning,this thesis researched the DF techniques based on neural network,respectively proposed the DF method based on Multilayer Perceptron(MLP)networks and based on Radial Basis Function(RBF)networks.Based on the Back Propagation(BP)algorithm,introduced the nonlinear conjugate gradient algorithm that can speed up the convergence speed.Then,the DF model based on MLP network is given and simulated.After that,introduced the K-means algorithm and Recursive Least Squares(RLS)estimation and gave the DF model based on RBF network,simulated this model under the same test conditions.Finally,focusing on the application background of the real-time DF when the semi-active seeker of anti-aircraft missile and airborne target relatively ultra-high-speed movement,and considering the jamming of Gaussian white noise and channel phase uncertainties,this thesis respectively proposed the DF scheme of missle-borne radar based on 32-element circular array and the DF scheme of missle-borne radar based on the optimal 10-element array.Meanwhile,the array optimization diagrams of the two schemes are given.Under the same test conditions,simulated and compared the two proposed DF schemes in all directions.