【作者】 李萍；

【导师】 郑翠娥；

【作者基本信息】 哈尔滨工程大学 ， 水声工程， 2020， 硕士

【摘要】 超短基线定位系统作为水下目标定位、导航等高精度水下作业的技术支撑,在海洋科学、国防工业等领域中发挥着日益重要的作用。为使水下目标可以得到足够的采样点数,需减小信号的发射周期以得到高数据刷新率,但当目标距离大于最大非模糊距离时,就会产生距离模糊问题,有效解决该问题是实现水下目标定位的重要技术之一。同时,由于水下环境的复杂性,时延差与斜距信息时常会受到野值的干扰,从而增大系统的定位误差。时延差和斜距信息的有效性也是高刷新率超短基线定位系统的重要研究内容,因此,在定位解算之前,需要采取有效手段对野值进行检测和修正。针对超短基线系统中的时延差与斜距数据出现野值的问题,本文研究一种基于自适应卡尔曼滤波的野值检测与修正算法。该算法将基于新息的野值检测方法与自适应模型相结合,采用衰减记忆滤波方法进行自适应噪声估计,增加野值检测地可靠性,同时,采用强跟踪滤波算法增加系统的鲁棒性。在野值修正上,针对正向卡尔曼滤波对斑点型野值修正精度不高的问题,采用正反向相结合的滤波法进行处理,最后采用RTS固定区间平滑算法提高野值修正精度。结果表明,与CV模型卡尔曼滤波相比,自适应抗野值算法能够有效地检测与修正野值。针对高刷新率超短基线定位系统存在的距离模糊问题,本文研究一种适用于超短基线的抗距离模糊方法:根据方向观测量与距离模糊无关这一特点,再由深度测量仪增加一个深度观测量,将方向观测量和深度观测量计算得到的斜距作为斜距粗值,将各个时刻的斜距粗值作为“初值装订法”的参考距离,解决了“初值装订法”先验初值难以获得的问题。最后,针对上述抗距离模糊方法在大基阵开角下不再适用的情况本文结合自适应卡尔曼滤波算法,通过仿真验证了该算法在野值检测与修正的同时,可以解决大基阵开角下的超短基线距离模糊问题。通过试验数据对本文所研究的算法进行验证,海试数据处理结果表明,基于卡尔曼滤波的自适应抗野值算法可以有效检测出野值,降低数据中的野点率;湖试数据处理结果表明,本文研究的抗距离模糊方法可以解决大基阵开角下的超短基线距离模糊问题,与斜距粗值计算得到的定位结果相比,该方法减小了超短基线的定位误差。更多还原

【Abstract】 Ultra-short baseline positioning system,capable of providing technical support for high-precision underwater operations,such as underwater target positioning and navigation,plays an increasingly important and prominent role in marine science,defense industry and other fields.The research on ultra-short baseline positioning systems focuses mainly on the validity of the delay difference and slope distance information which,due to the complexity of the underwater environment,are often disturbed by outliers,thereby increasing the positioning of the system error.Therefore,it is necessary to take effective measures to detect and correct outliers of delay difference and slope distance data.Besides,in order to ensure that sampling points for underwater targets are sufficient,the transmission period of the signal needs to be reduced to obtain a high data refresh rate,but the problem of distance blur will inevitably arise when the target distance is greater than the maximum non-blurred distance,thus contributing to resolving the distance.Particularly,a fuzzy problem serves as an important part in the research on the high refresh rate ultra-short baseline positioning system to achieve underwater positioning.To address the problems of outliers and skew data in ultra-short baseline systems,an outlier detection and correction algorithm are investigated in this paper based on adaptive Kalman filtering.Particularly,attenuation memory filtering method is adopted in this algorithm which combines innovation-based outlier detection method with adaptive model for adaptive noise estimation to make the outlier detection more reliable,along with the employment of a strong tracking filter algorithm to increase the robustness of the system.In the outlier correction,the forward and reverse filtering method,instead of the forward Kalman filter,is used to correct the speckle outlier.Finally,the RTS fixed interval smoothing algorithm is employed to improve the outlier correction accuracy.The results show that the adaptive anti-outlier algorithm,compared with the Kalman filter of the CV model,can help detect and correct outliers more effectively.To address the problem of distance ambiguity in the high refresh rate ultra-short baseline positioning system,an anti-distance ambiguity method suitable for ultra-short baselines is investigated in this paper,and a depth measuring instrument is conducted according to thefeature that the directional measurement has nothing to do with distance ambiguity.Depth view measurement,the slope distance calculated by the directional view measurement and depth view measurement,is employed as the rough distance value which is used as the reference distance of the "initial value binding method" each time,thus solving the "initial value binding method",making it difficult to obtain a priori initial value.Finally,in view of the fact that the above-mentioned anti-distance fuzzy method is no longer applicable to large base array opening angles,the fact that the algorithm can solve large base array opening and detection while outlier detection and correction is verified by simulation in this paper through combining with adaptive Kalman filtering algorithm.This helps address the problem of blurring the ultra-short baseline distance under the corner.The field algorithm is employed to verify the algorithm studied in this paper.The results of sea trial data processing show that outliers can be detected and the outlier rate in the data can be reduced effectively with the adaptive anti-outlier algorithm based on Kalman filtering.The results of lake test data processing demonstrate that the anti-distance blur method studied in this paper can address the problem of ultra-short baseline distance ambiguity at large base array opening angles,and can reduce the short baseline positioning error more effectively compared with the positioning results calculated by the rough value of the slope distance.更多还原