Application Of Flink In Multi-source Tracking Target Fusion System

With the prosperity and development of China’s shipping industry,the ship traffic in coastal and Yangtze River ports is very busy.In order to ensure the safety of ship traffic,China has introduced a large number of ship traffic management VTS(vessel traffic services)systems in recent 10 years.However,the security of ship traffic big data information and the data sharing and integration of regional VTS systems have become increasingly prominent,As a result,the localization process of VTS is accelerating.Multi source target fusion system is the core and key component of VTS.At present,all VTS systems adopt the traditional centralized fusion system,which can not meet the needs of large-scale VTS or regional VTS system.Based on the analysis and research of the existing VTS target fusion system in China,this thesis proposes to apply Flink streaming data processing technology with distributed characteristics to the multi-source target fusion system,so as to improve the calculation speed and reduce the system resource consumption,and separate the multi-source target fusion system from the VTS system and deploy it separately,So it can improve the efficiency of localization and application.The main research and design contents include:(1)This paper analyzes the data structure,transmission protocol and related algorithms of target fusion in the existing VTS system,studies the Flink software architecture,and establishes the application model of multi-source target fusion system based on Apache Kafka data stream transmission technology and redis data storage system.(2)This thesis studies the multi-source,mainly including the real-time detection information of various common types of radar and the dynamic data automatically reported by AIS ships.A track association algorithm based on multi factor fuzzy synthesis is proposed,and the association fusion is carried out according to the position,heading and speed information of the target tracking data.(3)The requirements of the system are analyzed and implemented.The tracking data of various sources are calibrated and normalized in time and space.The system is implemented by using open source programming platform.Through deployment and testing,the system is successfully integrated into the whole domestic VTS.