Research On Multi-rotor UAV SAR Imaging And Motion Compensation

Compared with the general fixed-wing aircraft SAR,multi-rotor UAV SAR has the advantages of flexible flight and convenience to use,and can play an important role in disaster relief and military reconnaissance.However,due to the unstable flight and difficult to load high-precision IMU,its motion error is difficult to compensate.Firstly,the echo model of multi-rotor UAV SAR including motion error is established,and the echo phase error caused by motion error is deduced in detail.Then,the Range-Doppler algorithms under side-looking mode and squint mode are derived respectively.Next,The effect of frequency and amplitude of track error on SAR imaging with different beam widths is analyzed.In the aspect of motion-parameter-based compensation,this paper divides the error into along-track motion error and line-of-sight motion error.These two errors are further subdivided into coordinate error and velocity error,which are analyzed and compensated respectively.In addition,this paper analyzes the limitations of two-step motion compensation and one-step motion compensation in detail,and puts forward Range-cell-by-cell Compensation method to solve the compensation problem when the track error’s amplitude is large and the error’s frequency is high.In the aspect of autofocus,the existing multi-subaperture autofocus algorithms are analyzed in this paper.Most of the existing multi-subaperture autofocus algorithms for strip-mode SAR rely on Phase Gradient Autofocus.Phase Gradient Autofocus has good effect only when there are strong scattering points in the scene and most of the errors have been corrected.However,strong scattering points don’t always exist,and because it is difficult for multi-rotor UAV to carry a high-precision IMU,the residual error of SAR echo data before autofocus may be large.Therefore,a multi-subaperture autofocus algorithm based on minimum entropy criterion is proposed.This algorithm is more robust than other multi-subaperture self-focusing algorithms.In this algorithm,when estimating the error of subapertures and calculating the shift amount of error gradients,relevant parameters are both searched based on the minimum entropy criterion.Finally,the robustness is verified by a simulation and a real data processing.When the error range-variability is strong,in order to obtain high-resolution SAR images,the compensation usually uses autofocus phase error to estimate the error in some range cells,and then use these error to restore the trajectory of the the radar antenna.However,the existing radar trajectory estimation methods are limited by the target distribution in the scene and only consider the inter-pulse error caused by the coordinate offset.Multi-rotor UAV SAR is generally FMCW SAR,which means that not only the inter-pulse error but also the in-pulse error needs to be compensated.Therefore,the existing methods for trajectory estimation can’t be directly applied to multi-rotor UAV SAR.This paper improves the existing trajectory estimation methods.The improved method solves the problem of uneven target distribution by dividing the whole aperture into multiple subapertures,and takes into account the coordinate offset and velocity error.It is combined with the Range-cell-by-cell Compensation method,and the motion track of SAR can be well restored through multiple iterations.A simulation verifies the accuracy of this multi-subaperture trajectory estimation method.