高频地波雷达与AIS船只目标航迹关联方法研究

发布时间：2018-01-15 22:27

本文关键词：高频地波雷达与AIS船只目标航迹关联方法研究　出处：《内蒙古大学》2014年硕士论文　论文类型：学位论文

【摘要】：在众多海上监视手段中,高频地波雷达能够对监视区域内的合作与非合作目标进行超视距探测和跟踪,但其不能提供目标的身份信息；舰船自动识别系统(Automatic Identification System, AIS)能提供监视区域内合作目标的信息,为目标的类型和身份识别提供了便利,但AIS不能提供非合作目标的信息。二者各有优缺点,提供的信息具有互补性,将二者得到的船只目标探测结果进行航迹关联,一方面可以解决监视区间中的重复跟踪问题,提高跟踪效率,另一方面可以通过排除关联的合作目标重点关注没有身份信息的疑似非合作船只。本文通过地波雷达与AIS船只目标航迹关联分析；发展适用于地波雷达与AIS的船只目标航迹关联算法,实现合作目标和非合作目标的跟踪探测。论文主要在以下几个方面做了研究工作： 1.在数据预处理方面,分析了地波雷达与AIS点迹与航迹数据的特点,并针对二者在时空上的非统一性开展了两种手段航迹的数据预处理。通过基于球体模型的坐标变换和径向投影变换,实现了坐标统一。采用分段线性插值和外推法对AIS航迹进行时间插值,实现了与地波雷达的时间对准。 2.在航迹关联算法方面,提出了基于均值最近邻的地波雷达与AIS船只目标航迹关联方法,并对门限参数的选取做了分析。该方法采用径向速度、距离和方向的三维特征信息,用一段时间内的均值最近邻判据进行航迹关联判定,该航迹关联方法比传统的最近邻法能够减小误判率。针对航迹较多的复杂情况,提出了模糊双门限航迹关联方法,并对该方法中参数的选取做了分析。将模糊双门限方法与均值最近邻法做了比较,在航迹较复杂或有奇异值的情况下,模糊双门限的方法具有较强的鲁棒性。基于对航迹关联结果的误差统计,提出了地波雷达航迹修正方法,提高了地波雷达对非合作船只的跟踪精度。 3.最后,在MATLAB环境下构建了地波雷达与AIS航迹关联软件系统；系统集成了均值最近邻和模糊双门限两种航迹关联算法和航迹修正方法,能够实现航迹关联和修正,可以突出显示非合作目标的原始及修正航迹。
[Abstract]:Among many marine surveillance methods, HF ground wave radar can detect and track the cooperative and non-cooperative targets in the surveillance area, but it can not provide the identification information of the target. Automatic Identification system (AIS) can provide information about cooperative targets in the monitoring area. It is convenient for the type and identification of the target, but AIS can not provide the information of the non-cooperative target. They have their own advantages and disadvantages, and the information provided is complementary. On the one hand, it can solve the problem of repeated tracking in the monitoring area and improve the tracking efficiency by correlating the ship target detection results obtained by the two methods. On the other hand, we can focus on the suspected non-cooperative ships without identity information by excluding the associated cooperation targets. In this paper, the correlation analysis between ground wave radar and AIS ship target track is carried out. The tracking and detection of cooperative and non-cooperative targets are realized by developing a ship target track association algorithm suitable for ground wave radar and AIS. In this paper, the following research work has been done: 1. In the aspect of data preprocessing, the characteristics of ground wave radar and AIS point track and track data are analyzed. In view of the non-unity of the two methods, the data preprocessing of track is carried out by means of coordinate transformation and radial projection transformation based on sphere model. The method of piecewise linear interpolation and extrapolation is used to interpolate the track of AIS, and the time alignment with ground wave radar is realized. 2. In the aspect of track correlation algorithm, a method of track correlation between ground wave radar and AIS ship target based on mean nearest neighbor is proposed, and the selection of threshold parameters is analyzed. The radial velocity is used in this method. The distance and direction 3D feature information is determined by the mean nearest neighbor criterion in a period of time. The track correlation method can reduce the misjudgment rate compared with the traditional nearest neighbor method. In view of the complex situation of more track, a fuzzy double threshold track correlation method is proposed, and the selection of parameters in the method is analyzed. The fuzzy double threshold method is compared with the mean nearest neighbor method. In the case of complex tracks or singular values, the fuzzy two-threshold method is robust. Based on the error statistics of track correlation results, the track correction method of ground wave radar is proposed, which improves the tracking accuracy of surface wave radar to non-cooperative ships. 3. Finally, the software system of track correlation between ground wave radar and AIS is constructed in MATLAB environment. The system integrates mean nearest neighbor and fuzzy double threshold algorithms and track correction methods, which can realize track correlation and correction, and can highlight the original and modified track of non-cooperative targets.
【学位授予单位】：内蒙古大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TN957.52

【参考文献】