Research Of Clutter Environment Perception And Radar Waveform Optimization Design

As an important means of comprehensive perception of the battlefield situation,radar has a pivotal position in modern warfare.With the increasingly complex battlefield environment,classical radar signal processing methods have been difficult to complete tasks such as target detection in a complex electromagnetic environment.In order to break the upper limit of the performance of classical radar signal processing and improve the performance of traditional radar systems such as detection,tracking and recognition,the concept of cognitive radar came into being.This paper uses cognitive radar as the research background,focusing on clutter environment perception and radar waveform optimization design methods,including the clutter environment perception based on deep learning methods and the ambiguity function shaping method based on Riemannian optimization,and combined experiment and analyze the measurement data of the typical clutter in microwave chamber.The main research work and innovations of this paper are summarized as follows:1.Aiming at the classic clutter perception problem,this paper studies the parameter fitting method based on Taylor expansion based on the Kulemin model and Ulaby model.This method first performs Taylor series expansion on the two classic electromagnetic scattering models,and then uses the existing public data set to perform parameter fitting on the expanded model,and then obtains the specific parameters of different clutter environments,and realizes the detection of typical ground objects.Effective perception of the environment.2.This paper proposes a clutter environment perception method based on deep learning and prototype network.Aiming at the problem of insufficient feature extraction of parameter fitting methods and inaccurate description results,this paper uses the idea of convolutional neural network and combines the measurement data of microwave anechoic chambers of typical objects to propose a data-driven environment identification method;The problem of small sample limitation caused by time observation window,this paper proposes a small sample clutter environment sensing method based on prototype network.This method uses the high-resolution one-dimensional range profile of the clutter environment as training and test data to achieve small samples Environmental identification of different polarization modes and wiping angles under different conditions.3.Aiming at the interference suppression problem of pulse Doppler radar near range loop,this paper proposes a radar waveform design method under the framework of Riemann optimization.This method converts the constant modulus constraint in Euclidean space into unconstrained Riemannian manifold Optimize the problem and solve it using an algorithm based on Riemann’s first-order and second-order steps.The simulation experiment results show that,compared with the classical optimization methods based on relaxation and approximation,the radar waveform solved by the method in this paper has better anti-jamming performance.