Data-driven detection and diagnosis on the navigation states of surface vessels

The surface vessel as a small surface warship capable of autonomous planning and navigation and mostly used to carry out various dangerous battle missions has been widely used by many countries in military field. First, the advantages and importance of surface vessels are introduced and the necessity of conducting fault detection and diagnosis on the navigation states is discussed as well. Second, the basic theories of three data-driven approaches are presented, including PCA based approach, SVM based approach and BPNN based approach. Third, we analyze the results of the three data-driven approaches and propose the accessible methods of detection and diagnosis, which provides references for researchers devoted to the related works. Finally, we discuss the challenges and prospects related to the work in this paper.