论文总字数：23840字

摘 要

随着船舶自动识别系统（Automatic Identification System，AIS）在船舶避碰，遇险报警，身份识别等方面的广泛应用，交互产生的船舶轨迹数据的规模以指数形式增长。海量数据的储存，在增加通信成本的同时会给终端数据库带来挑战：数据冗余，查询延时增长等。为应对上述挑战，多种轨迹压缩算法被提出。但是迄今为止缺少可以涵盖绝大多数船舶轨迹压缩算法的全面的对比和算法性能评估。为了对该领域涉及算法进行综述，本文的主要研究内容如下：

1.本文在收集归纳总结文献的基础上，对目前船舶轨迹数据领域的主流算法进行介绍。其次为了便于读者理解，本文通过结合图形的方式，举例对算法的逻辑和思想进行了阐述。

2.针对探究不同算法性能差异的问题，本文在Eclipse开发环境下使用java编程语言复现算法模型。并在实际生产场景下采集海量AIS数据样本，经过数据预处理得到可用数据。通过开展一系列对比实验，获取算法性能参考数据，进而对不同算法性能进行综合比对，获得对于船舶轨迹数据压缩算法性能的综合评价。

3.为了增强研究内容的实用性，本文提供了针对船舶轨迹数据离线和在线场景下的推荐压缩算法。该推荐算法在探究船舶轨迹数据压缩算法性能的综合评价的基础上，最大程度的满足船舶运营场景下对压缩算法精度和实时性的要求

研究结果表明：离线压缩算法中，DP算法与TD-TR算法综合性能相当，但后者对于轨迹数据压缩性能较为优异。在线压缩算法中，Uniform算法压缩效果因算法逻辑原因存在较大波动性，SQUISH算法和SQUISHE-E算法性能较好，但难以实现压缩效果和压缩误差的相对平衡。对于船舶轨迹数据在线压缩场景下，OPW-TR算法可以实现数据高效，精准压缩，进而消除数据冗余，降低数据传输通信开支，优化船舶自动识别系统性能。

关键词： 轨迹压缩；算法模型；误差度量；性能对比；算法推荐

ABSTRACT

With the wide application of Automatic Identification System (AIS) in collision avoidance, distress alarm, Identification and so on, the scale of ship trajectory data is increasing exponentially. The storage of data, while increasing the communication cost, will bring challenges to the Terminal Database: Data Redundancy, query delay growth and so on. To solve above challenges, many excellent trajectory compression algorithms have been proposed. But so far there is lack of comprehensive comparison and performance evaluation which can cover most of the ship trajectory compression algorithms. Therefore, in order to give a comprehensive overview of the algorithms involved in this field, the main research contents of this paper are as follows:

1. In this paper, the main ship trajectory compression algorithm is introduced in detail on the basis of literature collection and summary. In order to facilitate the reader to understand, this article through the combination of graphics, examples of the Algorithm on the logic and ideas.

2. In order to explore the performance difference of different algorithms, this paper uses Java programming language to reproduce the algorithm model in Eclipse development environment. A large number of AIS data samples are collected in the actual production scene, and the available data are obtained through the Data pre-processing. Through a series of comparison experiments, the reference data of the Algorithm performance is obtained, and then the performance of the different algorithms is compared, and the comprehensive evaluation of the performance of the ship trajectory data compression algorithm is obtained.

3. In order to enhance the practicability of the research, this paper provides a recommended compression algorithm for ship trajectory data in off-line and on-line scenarios. Based on the comprehensive evaluation of ship trajectory data compression algorithm, the proposed algorithm can meet the requirements of precision and real-time performance in ship operation

The results show that the overall performance of DP algorithm and TD-TR algorithm is comparable, but the latter is superior for trajectory data compression. Among the on-line compression algorithms, Uniform algorithm has great fluctuation due to its logic, SQUISH algorithm and SQUISHE-E algorithm have good performance, but it is difficult to achieve the relative balance between compression effect and compression error. In the case of ship trajectory data on-line compression, OPW-TR algorithm can compress data efficiently and accurately, thus eliminate data redundancy, reduce data transmission and communication expenses, and optimize the performance of ship automatic identification system.

Keywords：Trajectory compression; Algorithm Model; Error measurement; performance comparison; Algorithm recommendation

请支付后下载全文，论文总字数：23840字