Low Altitude Small Drone Radar Detection And Recognition

In recent years,small Unmanned Aerial Vehicle(UAV)is widely used in various fields,which brings great convenience to industrial production and daily life.However,the incidents that threaten the public safety caused by UAV "black fly" also occur frequently,which seriously interfere with the normal production and life order of people.In order to resist the security threats brought by UAV,the demand for its discovery,identification and countermeasures is also increasing.A s a typical "low slow small" target,UAV is difficult to be detected by radar.In addition to its small size,strong mobility,special structure,good at low altitude flight,this series of characteristics increase the difficulty of radar detection.Based on this,this paper studies the detection and recognition of UAV in low altitude environment.Firstly,based on the basic structure of the rotor UAV,the motion model of the UAV is established,and the characteristics of its flight process are studied.Bas ed on this model,the echo signal model of fuselage and rotor is derived,and the expression of echo signal is given,which lays the foundation for the follow-up work.Aiming at the low altitude multipath environment of UAV,this paper studies the detection performance of radar for different Swerling undulating targets in multipath environment based on the signal propagation model and target detection model in multipath environment.Through numerical simulation,a method for selecting parameters of M/N detector in multipath environment is given,and the effectiveness of the method is verified by simulation experiments.In order to improve the accuracy of parameter estimation of UAV in low signal-to-noise ratio scene,this paper proposes a parameter estimatio n method for hovering UAV based on the combination of inverse Radon transform and circle detect.It improves the existing parameter estimation method,improves the operation speed and anti noise performance.For mobile UAV,a parameter estimation method based on Radon transform and sinusoidal detection is proposed in this paper.After the motion compensation of mobile UAV,the method of parameter search is used to estimate its parameters.In addition,this paper also studies the application of deep learning in UAV target recognition.Combining the traditional signal processing method with convolutional neural network,this paper proposes a method of UAV recognition based on the combination of short-time Fourier transform and inverse Radon transform using Google Net.The recognition rate of this method is verified by the data set generated by simulation.