Abnormal Ship Behavior Detection Based on AIS Data

With the development of navigation globalization and ship dehumanization, contradiction between increasing demand for behavior supervision limited traffic service resources is obvious, frequent occurrence accidents at sea a problem. The monitoring abnormal an important link in maritime transportation. popularization automatic identification system research field, AIS widely used management static information real-time sharing dynamic information. generated moving trajectory data provide new opportunity into its detection. In light current situation research, we detected from point view spatial thematic based on data. Therefore, this study first modeled cognition behavior. Then, group rules, method graph structure learning to mine routes perspective Next, Rayda’s criterion detect anomalous ships space. isolation forest algorithm, described shown by experimental results show that framework proposed paper can effectively ships.