The Study of Fishing Vessel Behavior Identification Based on AIS Data: A Case Study of the East China Sea

The goal of this paper is to strengthen the supervision fishing behavior in East China Sea and effectively ensure sustainable development fishery resources. Based on AIS data, analyzes three types boats (purse seine operation, gill net operation trawl operation) uses cubic spline interpolation algorithm optimize ship trajectory construct high-dimensional features. It proposes a new coding method for boat sequences. This Geohash divide into grids generate corresponding numbers. Then, mapped grid, points are associated with divided ID extracted from grid. complete sequence passes through CBOW (continuous bag words) model, correlation fully learned. Finally, converted coordinate vector, processed trained by LightGBM algorithm. In order obtain optimal classification effect, superparameter combination selected. We put forward based Bayesian optimization algorithm, obtained results kinds boats. final result was evaluated using F1_score. Experimental show that F1_score proposed vectorization highest, training accuracy 0.925. Compared XgBoost CatBoost, increased 1.8% 1.2%, respectively. demonstrates strong applicability effectiveness area evaluations significant strengthening resource management.