Study On Parameter Estimation Algorithms Of Complex Targets For Large Stereo Arrays

With the rapid development of wireless eleconics technology and computer industry,human's life and production are constantly affected by electromagnetic waves,not only created huge wealth for human beings,but also improved many human lifestyles,but it has caused the electromagnetic environment to become more complicated and diversified.Therefore,more and more attention has been paid to related research in the electromagnetic field.As an important field of electromagnetic perception,direction of Arrival(DOA)estimation is a hot research topic in array signal processing.Improving the measurement and control capabilities of military equipment and accurately grasping the azimuth information of the target will lead the way to gaining the initiative in information warfare.Therefore,how to improve the DOA estimation algorithm's ability to deal with complex electromagnetic environments,improve the direction finding accuracy and angular resolution,and reduce the error rate in DOA estimation are difficult problems.Nowadays,the DOA estimation algorithms for linear and area arrays have continued to mature.The three-dimensional array expands conformer in three dimensions,which can cover the entire airspace more efficiently and has greater research significance.Aiming at the classic DOA estimation algorithms,the number of snapshots required is too large,the signal-to-noise ratio is high,and complex signal source such as coherence and strength in irregular formations direction finding problems are difficult to be solved,so the thesis study on high resolution and high precision DOA estimation agorithm for Stereo Array.First of all,this thesis introduces the ideas of three types of classic DOA estimation algorithms and a class of decoherent methods for solving regular arrays.Firstly,based on non-coherent signal sources,they compared their performance under different conditions through experimental simulations.Then,for the coherent signal source,the spatial smoothing method is adopted.The performances of the CAPON and MUSIC algorithms are compared through experimental simulations.The spatial smoothing method has many limitations in the separation of the coherent signal source.Subsequently,two types of algorithms based on compressive sensing theory were studied,including algorithms based on singular value decomposition and sparse Bayes algorithm.The sparse representation of array signals based on compressed sensing theory are introduced in detail.The compressed sensing signal model in the stereo array mode is derived,and the performances of the two types of algorithms are studied through experimental simulationAiming at the large stereo array,due to the particularity of its array structure,a new strategy for region division and region selection is proposed based on the uniform division of the sub-array and dimensionality reduction.The proposed strategy not only solves the problem of inconsistent received signals at different array element positions('shading effect'),but also achieves coverage of the entire airspace.This thesis is based on a certain type of large-scale vehicle conformal array,firstly,it compares the algorithm based on matrix singular value decomposition with the MUSIC algorithm that is the better performance of the traditional array of DOA estimation at the sub-array level.Finally,the algorithm based on matrix singular value decomposition is selected,which combines region division and region selection strategies to achieve multi-target complex signal source DOA estimation and can also achieve the high-precision and high-resolution direction finding for strong and weak signal sources or coherent signal sources,the effectiveness and feasibility of the algorithm is verified through simulation experiments.