Research On Marine Fishing Vessel Monitoring System And Related Technology Based On Edge Computing

Due to the complex and ever-changing marine environment,various fishery safety incidents occur from time to time.In order to protect the lives and property of fishermen,the supervision of fishing vessels is strengthened.A vessel monitoring system(VMS)that combines technologies such as satellite communications,global positioning and navigation,computers and geographic information has emerged.The existing VMSs all adopt a centralized centralized computing mode,and all visual display and pattern mining for fishing vessel data are completed in the monitoring center.However,due to the lack of maritime communication resources,the data collected by the terminal equipment cannot be utilized efficiently,and the communication delay causes the demand for high real-time performance to be unsatisfactory.As a fast-growing new computing model,edge computing can effectively solve the problem of high lagging and insufficient bandwidth by providing services close to the object or data source.Therefore,this dissertation applies the edge calculation mode to VMS,and on this basis,it studies the fishing vessel trajectory data transmission and anomaly detection.(1)This dissertation proposes a VMS(EC-VMS)framework based on edge computing.The framework is mainly composed of four parts: the sensing layer,the aggregation layer,the edge layer and the cloud layer.The sensing layer is responsible for collecting ocean information;the aggregation layer is responsible for filtering and integrating the collected data;the edge layer is the core layer of the EC-VMS framework,and tasks such as trajectory simplification and abnormality detection can be performed at the data source,and communication can be conducted with the help of Beidou navigation satellite system(BDS)and cloud layer;the cloud layer is responsible for storing fishing vessel information and providing services such as visual display and data analysis.(2)In view of the lack of maritime communication resources and the inability to efficiently utilize fishing vessel data,this dissertation proposes an adaptive trajectory transfer model(ATTM)based on the EC-VMS framework.The model utilizes the computing power of the edge nodes to establish an adaptive data transmission mechanism based on the satisfaction of the Beidou satellite communication.The LDR and SQUISH algorithms reduce redundant data and reduce the number of satellite communications.So that the packet loss feedback mechanism and error checking strategy ensure the reliability of data transmission.(3)Because the maritime communication delay can not effectively meet the realtime demand of the abnormal detection of fishing vessels,this dissertation proposes a real-time anomaly detection model(RADM)based on the EC-VMS framework.The model runs at the edge layer,and establishes a cooperation mechanism between the edge nodes,and combines the historical trajectory extraction detection algorithm with the online anomaly detection algorithm to realize the anomaly detection function.The historical trajectory extraction detection algorithm mines frequent patterns in historical trajectories by multi-feature clustering,and identifies trajectories different from frequent patterns as anomalies;online anomaly detection algorithm detects abnormal behaviors in specific scenarios based on spatio-temporal neighbor similarity and reduces the impact of abnormal evolution.