【作者】 孙伟；

【导师】 李海森；

【作者基本信息】 哈尔滨工程大学 ， 水声工程， 2015， 硕士

【摘要】 随着人类社会不断的发展,对于海洋资源的需求也越来越大,合成孔径声纳作为一种高效的海洋探测工具,能够为海洋资源的开发提供有力的技术支持。传统的侧扫合成孔径声纳,在声纳的正下方区域地距分辨率很低,会形成测绘盲区,多波束合成孔径声纳结合了合成孔径技术和多波束测深技术,解决了这一问题,实现了测绘区域的全覆盖、无盲区测量,具有很好的发展前景。由于海洋环境的影响,多波束合成孔径声纳的载体会偏离理想的运动轨迹,从而影响多波束SAS的成像质量,因此运动估计与补偿技术是提高多波束SAS成像质量的有效手段。成像效果是检验运动估计算法有效性的重要依据,而回波模拟是成像算法研究的基础,基于此本文的主要内容包括回波数据模拟、成像算法、运动估计技术三个方面。首先从合成孔径技术和多波束测深技术出发,研究了多波束SAS的基本原理。合成孔径声纳方位向分辨率高,但是存在测绘盲区。多波束测深声纳的方位向分辨能力较低,会随着测绘距离、信号波长而变化,但是可以进行全覆盖的测量。多波束SAS集合了二者的优势,是一种具有高方位向分辨率的三维成像声纳。然后分析了合成孔径声纳理论分析中常用的停走停假设和等效相位中心假设带来的误差,研究结果表明停走停假设带来的误差不可忽略,而等效相位中心假设对于回波信号的相位影响较小,因此论文的研究采用了等效相位中心假设下的匀速运动模型。在目标三维回波数据的模拟中,考虑了多波束SAS的运动模型、水声传播模型、水下三维地形生成和水下目标阴影特性等因素。接下来分别研究了 LFM信号和CW信号的成像算法,在方位向合成算法中,采用了最精确、最直接的时域BP成像算法,为运动补偿部分提供支撑。最后针对常规的多波束SAS的成像分辨率受限于预设波束数目的问题,提出了多波束SAS相干成像算法,该算法可以在不增加硬件复杂度的前提下,提高了成像质量。在多波束SAS载体运动估计研究中,首先分析了载体的六个自由度的运动误差对多波束SAS成像质量的影响,然后对成像质量影响较大的侧摆、航偏和横滚分别进行了估计与补偿,并分别进行了计算机仿真验证了算法的有效性。最后针对已有运动估计算法的不足,提出了一种六自由度联合估计的算法,这种算法考虑到载体的所有运动的可能性,具有更高的估计准确性。更多还原

【Abstract】 As the continuous development of human society,the demand for marine resources is also growing. Synthetic aperture sonar is an efficient marine exploration tool that can provide strong technical support for the development of marine resources. When projecting the sonar returns onto the seafloor, ground range resolution deteriorates as range decreases. Any side-looking sonar will therefore have a gap around nadir, or at the very least, a sector with severely degraded performance. While the multibeam synthetic aperture sonar technology combines synthetic aperture sonar and multibeam bathymetric technology to achieve full coverage mapping area. It is no blind measurements with good development prospect.Due to the impact of the marine environment, multibeam synthetic aperture sonar carrier will deviate from the ideal trajectory, thus affecting the image quality, therefore the motion estimation and compensation techniques are effective means to improve the image quality of multibeam SAS. Imaging results are an important basis to test whether the motion estimation algorithm is effective or not, and the echo simulation algorithm is the basis of imaging studies.So the main contents of this paper contain three sections: echo simulation, imaging algorithm and motion estimation.The basic principle of multibeam SAS is studied from the SAS and multibeam bathymetric technology. Comparing the azimuth resolution of the real aperture with synthetic aperture sonar, discussing the selection of the pulse repetition frequency. The azimuth resolution of the multibeam bathymetric sonar changes with the mapping distance increases and the signal wave length, while the azimuth resolution of multibeam SAS is fixed, which only determined by the element aperture length. Finally, considering the velocity, hardware complexity and other factors,a multibeam SAS of multi-array structure is given.In echo data simulation, the error of the Hop-and-Stop and the displaced phase center is studied, the results show the error caused by the Hop-and-Stop can’t be ignored, so we adopt the uniform motion models. In somatic target echo model, the point of the surface is used to calculate the echo delay and the four corners of the target is used to calculate the shadow area.Then the imaging algorithms of LFM signal and CW signals is studied. In azimuth synthesis algorithm, the paper adopt the most direct and the most accurate time-domain BP imaging algorithm to provide the support of the motion compensation. In order to solve the problem of the bathymetric resolution is limited by the beam number in the tradition multibeam SAS,a new multibeam SAS interferometric imaging algorithm is proposed.In the motion estimation of multibeam SAS section, firstly the paper analysis the impact of motion error for multibeam SAS image quality. The sonar platform has 6-DOF, Sway and Heave have great influence on the slant range of the target, the Roll impact on the beamforming across-track, the Yaw and the Pitch change the zone of sonar coverage. Then the Sway、Yaw and Roll were carried out estimation and compensation, the computer simulation results verify the effectiveness of the algorithm. Finally a new motion estimation algorithm of 6-DOF was proposed for the higher estimation accuracy, but it requires at least three non-collinear target in the imaging region.更多还原