Research And Implementation Of SAS Signal Imaging And Motion Compensation Technology

Synthetic Aperture Sonar(SAS)is a high-resolution imaging sonar.The basic principle is to use the uniform motion of a small aperture array to obtain a virtual large aperture in the azimuth direction,thus obtaining a high resolution in the azimuth direction.Unmanned Underwater Vehicle(UUV)is an underwater platform capable of autonomous navigation and intelligent operation.It is not only stable in operation,but also carries sensor devices and is convenient for motion compensation.It is an ideal carrier for SAS.SAS onboard UUV has broad potential in both civil and military application,and is one of the research hotspots in the field of SAS.This dissertation relies on enterprise cooperation projects to research SAS system signal processing technology onboard UUV,including motion compensation,imaging algorithms,signal processing system implementation and so on.The degree of freedom error of UUV on the three-dimensional carrier coordinate system has a great influence on the quality of SAS imaging,motion compensation technology is the most direct and effective way to eliminate or reduce motion error.The research on motion compensation mainly includes two aspects:(1)For the position information of the backhaul measurement system of UUV carrier,a SAS motion compensation method based on time delay table is proposed to avoid the need for track change.Repeatedly update the amount of computation that the delay table brings.(2)Analyze the phase error caused by the "stop-and-hop" hypothesis and the "displaced phase center" approximation,and propose an accurate multi-receiving array SAS compensation method,and pass simulation experiments,pool experiments and lake test experiments.The imaging results verify the effectiveness of the method.In terms of imaging algorithms,this dissertation proposes an improved imaging method for the shortcomings of conventional time domain algorithms,and the improved computational complexity is greatly reduced.At the same time,the influence of azimuth velocity disturbance on SAS imaging quality is analyzed.A simple and efficient speed smoothing method is proposed to reduce the loss of SAS image quality caused by the measurement error of inertial measurement system.In the aspect of signal processing system implementation,this dissertation designs a multi-thread parallel processing structure,which divides the whole signal processing flow into 5 threads for parallel processing,and can turn on the GPU(image processing unit)to accelerate the operation if necessary.For multi-batch data processing,a pipelined mode is used to achieve an efficient highresolution imaging system.It is verified by the pool experiment that the system works stably and the motion compensation effect is obvious.The imaging resolution can reach 5cm×2cm,and the maximum unambiguous detection distance can reach 180 m.