Page 109 - Demo

7Automatic Identification System Data Fusion97Marine Research of the National Academy of Marine ResearchDOI: 10.29677/MR.202406_4(1).0006 This study presents the development of an operational system for data quality management in theAutomatic Identification System (AIS). The system focuses on analyzing and debugging decoded AIS static and dynamic data. It systematically analyzes different ship behaviors under various navigation states and infers data and behavior patterns of ship AIS data based on these characteristics. Through the analysis of AIS data,the system automatically identifies abnormal ship data and removes these abnormal data points. The filtered data is then stored in a database for use in a dynamic ship information system, providing real-time and historical AIS data. It is noteworthy that the performance and real-time nature of AIS data largely depend on the stability and update speed provided by the data source. The AIS data used in this study was provided by the NAMR. During the process of AIS data quality management and abnormal data identification, a considerable amount ofabnormal data was discovered to be linked to repetitive AIS data. This result is inferred to be due to the receptionof AIS data from the same vessel by adjacent AIS base stations, resulting in a large amount of repetitive AISdata. Incomplete codes or incorrect ship MMSI codes also account for a larger percentage of abnormal data errors. As mentioned earlier, we have explained that the correct format for MMSI codes should be 9 digits, and according to specifications of ITU-R M.585-8 and ITU-T E.217 recommendations, different purposes have different MMSI assignment categories. Therefore, our assessment of abnormal data related to MMSI may be due to AIS base stations receiving AIS data from maritime electronic buoys, handheld VHF transceivers, and other radio beacons. The main reason is that, currently, we cannot verify whether the received AIS data istransmitted by onboard AIS equipment. For future studies, the ability to independently receive and decode AIS signals would be beneficial. This would significantly enhance the accuracy and breadth of ship behavior analysisin this system, enabling more comprehensive and precise identification of abnormal AIS data.5 CONCLUSIONS This article is part of the research project supported by the National Academy of Marine Research, No. NAMR111062. The author thanks the National Academy of Marine Research for funding support.ACKNOWLEDGEMENTS