The study of identification of fishing vessel behavior based on VMS data

Accurate identification of different behaviours of fishing vessels is important for fisheries management and fisheries ecology, which can enhance the management of overfishing and marine resources. In this study we identify fishing vessel fishing behavior through Vessel Monitoring System (VMS) data and BP neural networks. The change trend of the direction angle and speed of the fishing vessel is selected as the input parameters of the model, and the accuracy of identifying the fishing behavior is 79%. The fishery distribution is drawn according to the behavior of the fishing vessel identified by the model, which is similar to the distribution of the actual fishery and the fishing density. This laid the foundation for the deep exploration of the spatio-temporal characteristics of VMS in the future and the high-precision prediction of the distribution of fishing areas in China's offshore fisheries.