Identifying ship’s abnormal behavior by using principal component analysis and support vector machine in combination

In order to solve the problem of low recognition rate of abnormal behavior of ships, a method of identifying abnormal behavior of ships based on Principal Components Analysis（PCA） and Support Vector Machine（SVM） is proposed, which comprehensively considers ship attributes and motion characteristics. The ship behavior patterns are divided by considering the ship operation process, and the ship motion characteristics related to the ship behavior patterns are defined; Then, based on Automatic Identification System（AIS） data, PCA algorithm is used to extract the most representative ship motion features. Finally, SVM algorithm is used to identify the abnormal behavior of ships. The ship trajectory in the waters near Qingdao Port is selected for experimental analysis. The results show that there are 10 most representative ship motion features in the waters near Qingdao Port, with a total contribution rate of 84.40%. The average accuracy rate and average recall rate of ship abnormal behavior identification results are 83% and 84%, respectively, which are better than the comparison model. The research results can provide support for intelligent discovery and maritime supervision of abnormal situations such as abnormal ship position, abnormal trajectory, abnormal speed and abnormal course.