论文总字数：38899字

摘 要

随着自动识别系统(Automatic identification system, AIS)的强制使用，以及AIS数据的快速发展与广泛应用，使得监管船舶安全航行与水上交通安全的研究已经越来越依赖于船舶AIS轨迹数据。而AIS数据存在着大量无效和冗余信息，这增加了计算机存储成本，降低了数据调用效率。因此，如何在保留AIS轨迹主要结构特征的基础上压缩无效和冗余的信息具有重要的研究意义。

因此，本文以船舶AIS轨迹数据为研究对象，数据压缩技术为方法，针对原始开放窗口算法(Open Window, OPW)压缩过程中存在的相关问题，构造了一种基于OPW算法的改进窗口算法，来处理在线AIS数据的压缩。提高压缩后AIS轨迹点对于原轨迹的信息保留度，同时提升一定的压缩率来方便计算机的存储和调用。本文的主要工作如下：

(1) AIS数据的采集、解码和处理。在对AIS的信息种类、更新的频率进行介绍基础上，分析了原始AIS信息的主要构成，以及压缩过程中所产生的异常数据和噪声数据种类，并对AIS信息进行解码，对异常数据进行了剔除。提出了解决残缺的AIS信息的插值方法，以获得完成的AIS信息。

(2)为了验证改进的压缩算法的有效性，在原有OPW算法的基础上，引入了另外四种传统轨迹压缩算法。并就原始OPW算法原理和特点进行探究，提出了改进开窗算法模型，并演示了算法压缩过程。此外，还介绍了实验相关数据来源以及AIS数据量的大小。

(3)根据所引入的评价指标，对六种算法进行了实验对比，研究了在线、离线压缩算法的特性，为研究AIS轨迹数据压缩提供了新的依据。并对实验结果进行可视化研究来定性分析改进算法的有效性。

论文主要研究了改进的开窗算法在在线压缩中对船舶轨迹数据压缩的有效性。

研究结果表明：改进的开窗算法在高强度压缩的情况下压缩效果接近DP算法。

本文的特色：

1. 详细介绍了在海上领域有关AIS研究的多个方向。
2. 分别展示了国内外有关AIS在线压缩研究的进展。
3. 实验数据源较为丰富，数据量较大，为最终实验的有效性提供了数据支撑。
4. 引入DTW距离和距离损失率作为新的评价指标。
5. 最后依据可视化分析结果，来直观的评估改进OPW算法的有效性。

关键词：船舶轨迹；自动识别系统(AIS)；轨迹压缩；在线压缩；开窗算法

Abstract

With the mandatory use of Automatic identification system (AIS) and the rapid development and wide application of AIS data, the research on the supervision of safe navigation of ships and water traffic safety has become more and more dependent on the AIS track data of ships. AIS data has a lot of invalid and redundant information, which increases the cost of computer storage and reduces the efficiency of data invocation. Therefore, it is of great significance to research how to compress invalid and redundant information on the basis of retaining the main structural features of the AIS locus.

Therefore, this paper takes ship AIS trajectory data as the research object and data compression technology as the method. To solve the related problems in the original open window algorithm (OPW) compression process, an improved window algorithm based on OPW algorithm is constructed to deal with online AIS Data compression. Improve the information retention of the original trajectory of the AIS track point after compression, and increase a certain compression rate to facilitate the storage and recall of the computer. The main work of this article is as follows:

(1) AIS data collection, decoding and processing. Based on the introduction of the types of AIS information and the frequency of updates, the main components of the original AIS information, as well as the types of abnormal data and noise data generated during the compression process are analyzed, the AIS information is decoded, and the abnormal data Eliminate. An interpolation method to solve the incomplete AIS information is proposed to obtain the completed AIS information.

(2) In order to verify the effectiveness of the improved compression algorithm, four other traditional trajectory compression algorithms are introduced based on the original OPW algorithm. The principle and characteristics of the original OPW algorithm are explored, an improved windowing algorithm model is proposed, and the algorithm compression process is demonstrated. In addition, the data sources related to the experiment and the size of the AIS data are also introduced.

(3) According to the evaluation indexes introduced, six algorithms were compared experimentally, and the characteristics of online and offline compression algorithms were studied， providing a new basis for the study of AIS track data compression. The effectiveness of the improved algorithm is qualitatively analyzed by visual study of the experimental results.

请支付后下载全文，论文总字数：38899字