| The HFSWR(High Frequency Surface Wave Radar,HFSWR)of motion platform not only contains most of the functions of fixed platform HFSWR,but also has better flexibility and mobility.It also extends the detection range of HFSWR and can complete the detection tasks in designated sea areas,which has high application value in marine rights protection,law enforcement of maritime police,sea state monitoring,etc.When the HFSWR of motion platform is used for target detection,due to the movement of the plarform,the spectrum of the first-order sea clutter with strong energy will be broadened,thereby annihilating the ship targets with low radial velocity,and seriously affecting the target detection performance.Based on the existing sea clutter suppression methods and measured data,this dissertation will conduct an in-depth study on the sea clutter spectrum spreading mechanism,echo model,multi-dimensional characteristic analysis,and clutter suppression methods.The statistical characteristics of the sea clutter in the angle-Doppler domain are used and effective clutter suppression methods are proposed.This dissertation provides a feasible solution to the problem of low radial velocity target detection in the HFSW R system of motion platform.The main research contents of this dissertation are as follows:(1)The power spectrum of sea clutter echo is analyzed with the influence of platform movement.First,the mathematical models of three types of motion conditions,namely uniform linear motion,variable speed linear motion and six degrees-of-freedom(DOF)oscillatory motion,are established.Secondly,taking the sea clutter scattering cross sections as the research object,based on the electromagnetic scattering theory of rough sea surface,the first-order and secondorder power spectrum models under the above three types of motion conditions are established.Finally,the influence of the platform’s forward linear motion and 6 DOF oscillatory motion on the power spectrum of sea clutter echo is simulated.This part lays a theoretical foundation for the subsequent establishment of the space-time spectrum model of sea clutter and the analysis of the space-time characteristics.(2)From the perspective of signal processing,the space-time two-dimensional model of the echo signal in HFSWR of motion platform is established,and the influence of the platform motion on the space-time spectrum of both the target echo and sea clutter echo is theoretically deduced.The space-time spectrum characteristics of the echo in the case of the forward linear motion and six degrees of freedom oscillatory motion are simulated and analyzed.Aiming at eliminating the influence of the platform’s variable-speed motion and six-degree-of-freedom oscillatory motion on the echo,a motion compensation method based on the motion parameters is proposed.After compensation,the signal-to-noise ratio of the target is improved,and the linear space-time coupling relationship of sea clutter is recovered.The effectiveness of the method is verified through simulation experiments.Based on the measured data,the homogeneity of the echo data in the range dimension is analyzed by using the subspace-based range dimension correlation analysis method,which provides theoretical guidance for the designing of clutter suppression methods for limited training samples.(3)When the platform motion parameters and radar system parameters are determined,the echo signal can be compensated to obtain the sea clutter space-time spectrum with linear coupling characteristics.Utlizing the space-time distribution of sea clutter,two space-time adaptive processing(STAP)algorithms based on this priori knowledge are proposed,which are the auxiliary channel STAP algorithm and the sample transformation STAP algorithm.The auxiliary channel STAP adopts the indirect STAP algorithm structure according to the minimum mean square error criterion,uses the space-time channel on the clutter ridge as the virtual auxiliary channels to estimate the clutter component in the space-time channel to be detected,and then realize clutter suppression through cancellation.Since this method only selects the auxiliary channel on the clutter ridge,when the prior information of the space-time distribution of sea clutter is unstable,the performance of the algorithm is affected,and the robustness is poor.The sample transformation STAP algorithm adopts a direct structure according to the linear constraint minimum variance(LCMV)criterion,it selects training samples from the clutter ridge and range domain and removes the spatial frequency and time frequency difference between the training sample and the local area to be processed through phase transformation.The samples from the clutter ridge and different range cells form a new training sample set and the clutter covariance matrix is estimated using this new training sample set.This algorithm increases the sea clutter information in the range domain and improves the robustness of the algorithm.Experiments with measured data verify the effectiveness of these two algorithms.(4)Aiming at the problem of low availability of training samples in the range dimension caused by complex sea conditions,three STAP algorithms for solving lowavailability training samples are proposed.For the situation that some training samples have poor correlation(the number of training samples with a correlation coefficient greater than 0.7 is more than 50% of the number required by the Reed–Mallett–Brennan(RMB)criterion),a sample correlation weighted STAP algorithm is proposed.The correlation coefficients between the samples and the data to be processed are used to weight the clutter covariance matrix during calculation,which improves the accuracy of the CCM estimation.For the situation that most training samples are poorly correlated(the number of training samples with a correlation coefficient greater than 0.7 is more than 20% and less than 50% of the number required by the RMB criterion),a method of predicting the samples through Unscented Transformation(UT)to achieve sample expansion is proposed,and a more accurate estimation of the training sample covariance matrix is obtained.By summing and averaging the covariance matrix of the training samples,the estimation accuracy of the CCM in the range unit to be processed is improved.In the extreme situation that few training samples are available(the number of training samples with a correlation coefficient greater than 0.7 is less than 20% of the number required by the RMB criterion),combined with the sparse characteristics of the sea clutter at a specific frequency,a sparse reconstruction STAP algorithm based on the Sparse Bayesian Learning(SBL)algorithm is proposed.The sparse reconstruction of the CCM realizes the effective suppression of sea clutter under the condition of one single training sample in range dimension.The performance of the proposed clutter suppression algorithms is evaluated by the measured data and results show that the proposed methods have better performance on clutter suppression.The research results of this dissertation help to improve the clutter suppression performance of the motion platform HFSWR,which is of great significance to the application in target detection. |
PDF Full Text Request