Developing an Artificial Intelligence-Based Method for Predicting the Trajectory of Surface Drifting Buoys Using a Hybrid Multi-Layer Neural Network Model

Miaomiao Song,Wei Hu,Shixuan Liu,Shizhe Chen,Xiao Fu,Jiming Zhang,Wenqing Li,Yuzhe Xu

DOI: https://doi.org/10.3390/jmse12060958

IF: 2.744

2024-06-08

Journal of Marine Science and Engineering

Abstract:Accurately predicting the long-term trajectory of a surface drifting buoy (SDB) is challenging. This paper proposes a promising solution to the SDB trajectory prediction based on artificial intelligence (AI) technologies. Initially, a scalable mathematical model for trajectory prediction is developed, transforming the challenge of predicting trajectory points into predicting velocities in eastward and northward directions. Subsequently, a four-layer trajectory prediction calculation framework (FLTPCF) is established, outlining a complete workflow for the real-time online training of marine environment data and SDBs' trajectory prediction. Thirdly, for facilitating accurate long-term trajectory prediction, a hybrid artificial neural network trajectory prediction model, named CNN–BiGRU–Attention, integrates a Convolutional Neural Network (CNN), Bidirectional Gated Recurrent Unit (BiGRU), and Attention mechanism (AM), tuned for spatiotemporal feature extraction and extended time-series reasoning. Extensive experiments, including ablation studies, comparative analyses with state-of-the-art models like BiLSTM and Transformer, evaluations against numerical methods, and adaptability tests, were conducted for justifying the CNN–BiGRU–Attention model. The results highlight the CNN–BiGRU–Attention model's excellent convergence, accuracy, and generalization capabilities in predicting 24, 48, and 72 h trajectories for SDBs with varying drogue statuses and under different sea conditions. This work has great potential to promote the intelligent degree of marine environmental monitoring.

oceanography,engineering, marine, ocean

What problem does this paper attempt to address?

The problem that this paper attempts to solve is to accurately predict the long - term trajectories of Surface Drifting Buoys (SDBs). Specifically, the main challenges faced by the researchers include:

1. **Constructing a mathematical model**: How to construct an SDB trajectory prediction mathematical model that can seamlessly convert the trajectory point prediction problem into a velocity prediction problem, and ensure that both scalar and vector ocean environmental data can be integrated into the model to participate in trajectory calculations.
2. **Learning spatio - temporal features**: How to automatically learn the spatio - temporal variation features of ocean environmental factors and the historical trajectory information of SDBs to adapt to the changes in SDB structures and different flow conditions such as gentle currents, eddies and turbulences.
3. **Integrating neural networks**: How to organically integrate hybrid neural network models (such as CNN, GRU and DNN) and finely adjust the hyper - parameters to improve the accuracy of SDB trajectory prediction.

To address these challenges, the paper proposes an artificial intelligence - based method, especially by developing a hybrid multi - layer neural network model (CNN - BiGRU - Attention) to predict the trajectories of SDBs. This model combines Convolutional Neural Network (CNN), Bidirectional Gated Recurrent Unit (BiGRU) and Attention Mechanism, aiming to extract spatio - temporal features and perform long - time - series reasoning. Through a series of experiments, the excellent convergence, accuracy and generalization ability of this model in predicting the 24 - hour, 48 - hour and 72 - hour trajectories of SDBs with different towing states under different sea conditions are verified. This work is expected to enhance the intelligent level of ocean environmental monitoring.