# Automatic Identification System (AIS) Data Supported Ship Trajectory Prediction and Analysis via a Deep Learning Model

^{1}

^{2}

^{3}

^{4}

^{5}

^{6}

^{*}

## Abstract

**:**

## 1. Introduction

## 2. Methodology

#### 2.1. AIS Data Normalization

#### 2.2. AIS Data Prediction with the Bi-LSTM Model

#### 2.2.1. Basic LSTM Model

#### 2.2.2. Ship Trajectory Prediction with the Bi-LSTM Model

#### 2.3. Evaluation Metrics

## 3. Experiment

#### 3.1. Experimental Setups

^{−4}. We set the batch size into 64, and the MSE function is introduced as the loss function in the proposed deep learning framework. It is observed that model performance is also affected by network parameter settings, while input feature number and hidden layer node number play a more important role. The optimal settings for the input feature number and hidden layer node number for different models (i.e., Bi-LSTM, LSTM, SOC-LSTM and SOC-GAN) are shown as Table 3. Note that we did not provide the normalized AIS data in the following sections for the purpose of better readability.

#### 3.2. Single-Ship Trajectory Prediction and Analysis

^{−1}, which was obviously larger than the counterparts of Bi-LSTM, LSTM and SOC-LSTM. The MAPE indicator variation tendency showed similar performance compared to those of the MAE indicator. More specifically, the MAPE for the Bi-LSTM was 6.74 × 10

^{−5}, which was about half to that of the LSTM model (i.e., 1.15 × 10

^{−4}).

^{−4}and 4.13 × 10

^{−4}, respectively. We found that the MSE values were smaller than those of the MAE and MAPE. It was observed that MSE for the Bi-LSTM model was 1.43 × 10

^{−9}, which was significantly lower than the counterparts of the MAE and MAPE. The MSE values for the LSTM, SOC-LSTM and SOC-GAN were 1.94 × 10

^{−9}, 4.63 × 10

^{−7}and 7.79 × 10

^{−2}, which were at least 30% larger than that of the Bi-LSTM model. In sum, the Bi-LSTM model can accurately predict ship trajectory in terms of the MAE, MSE and MAPE indicators compare to those of the LSTM, SOC-LSTM and SOC-GAN models.

#### 3.3. Trajectory Prediction for Multiple Ships

^{−5}, 9.07 × 10

^{−5}and 3.98 × 10

^{−9}, which were smaller than the counterparts of other three models. It was noted that the MAE indicator for the LSTM model was approximately three-fold larger than that of the Bi-LSTM model. The MAPE and MSE indicators of the LSTM model, which were 1.31 × 10

^{−4}and 1.02 × 10

^{−8}, respectively, showed similar performance compared to that of the MAE indicator. The aggregated MAE, MAPE and MSE for the SOC-LSTM were at least two-fold larger than those of the LSTM model, which were 3.57 × 10

^{−4}, 7.98 × 10

^{−4}and 3.31 × 10

^{−7}. The three indicators (i.e., MAE, MAPE and MSE) for the SOC-GAN, shown in the Table 7, were smaller than those of the SOC-LSTM.

## 4. Conclusions

## Author Contributions

## Funding

## Institutional Review Board Statement

## Informed Consent Statement

## Conflicts of Interest

## References

- Benz, L.; Münch, C.; Hartmann, E. Development of a search and rescue framework for maritime freight shipping in the Arctic. Transp. Res. Part A Policy Pract.
**2021**, 152, 54–69. [Google Scholar] [CrossRef] - Zhou, Y.; Daamen, W.; Vellinga, T.; Hoogendoorn, S. Review of maritime traffic models from vessel behavior modeling perspective. Transp. Res. Part C Emerg. Technol.
**2019**, 105, 323–345. [Google Scholar] [CrossRef] - Ma, D.; Ma, W.; Hao, S.; Jin, S.; Qu, F. Ship’s response to low-sulfur regulations: From the perspective of route, speed and refueling strategy. Comput. Ind. Eng.
**2021**, 155, 107140. [Google Scholar] [CrossRef] - Özlem, Ş.; Altan, Y.C.; Otay, E.N.; Or, İ. Grounding probability in narrow waterways. J. Navig.
**2020**, 73, 267–281. [Google Scholar] [CrossRef] - Gao, M.; Shi, G.-Y. Ship-handling behavior pattern recognition using AIS sub-trajectory clustering analysis based on the T-SNE and spectral clustering algorithms. Ocean Eng.
**2020**, 205, 106919. [Google Scholar] [CrossRef] - Rawson, A.; Brito, M. A critique of the use of domain analysis for spatial collision risk assessment. Ocean Eng.
**2021**, 219, 108259. [Google Scholar] [CrossRef] - Ma, D.; Hao, S.; Ma, W.; Zheng, H.; Xu, X. An optimal control-based path planning method for unmanned surface vehicles in complex environments. Ocean Eng.
**2022**, 245, 110532. [Google Scholar] [CrossRef] - Rodger, M.; Guida, R. Classification-aided SAR and AIS data fusion for space-based maritime surveillance. Remote Sens.
**2020**, 13, 104. [Google Scholar] [CrossRef] - Rong, H.; Teixeira, A.; Soares, C.G. Maritime traffic probabilistic prediction based on ship motion pattern extraction. Reliab. Eng. Syst. Saf.
**2022**, 217, 108061. [Google Scholar] [CrossRef] - Dalsnes, B.R.; Hexeberg, S.; Flåten, A.L.; Eriksen, B.-O.H.; Brekke, E.F. The neighbor course distribution method with Gaussian mixture models for AIS-based vessel trajectory prediction. In Proceedings of the 2018 21st International Conference on Information Fusion (FUSION), Cambridge, UK, 10–13 July 2018; pp. 580–587. [Google Scholar]
- Zhang, C.; Bin, J.; Wang, W.; Peng, X.; Wang, R.; Halldearn, R.; Liu, Z. AIS data driven general vessel destination prediction: A random forest based approach. Transp. Res. Part C Emerg. Technol.
**2020**, 118, 102729. [Google Scholar] [CrossRef] - Murray, B.; Perera, L.P. A dual linear autoencoder approach for vessel trajectory prediction using historical AIS data. Ocean Eng.
**2020**, 209, 107478. [Google Scholar] [CrossRef] - Herrero, D.A.; Pedroche, D.S.; Herrero, J.G.; López, J.M.M. AIS trajectory classification based on IMM data. In Proceedings of the 2019 22th International Conference on Information Fusion (FUSION), Ottawa, ON, Canada, 2–5 July 2019; pp. 1–8. [Google Scholar]
- Zhang, Z.-G.; Yin, J.-C.; Wang, N.-N.; Hui, Z.-G. Vessel traffic flow analysis and prediction by an improved PSO-BP mechanism based on AIS data. Evol. Syst.
**2019**, 10, 397–407. [Google Scholar] [CrossRef] - Yu, Q.; Liu, K.; Teixeira, A.P.; Soares, C.G. Assessment of the Influence of Offshore Wind Farms on Ship Traffic Flow Based on AIS Data. J. Navig.
**2020**, 73, 131–148. [Google Scholar] [CrossRef] - Liu, C.; Liu, J.; Zhou, X.; Zhao, Z.; Wan, C.; Liu, Z. AIS data-driven approach to estimate navigable capacity of busy waterways focusing on ships entering and leaving port. Ocean Eng.
**2020**, 218, 108215. [Google Scholar] [CrossRef] - Chen, X.; Ling, J.; Wang, S.; Yang, Y.; Luo, L.; Yan, Y. Ship detection from coastal surveillance videos via an ensemble Canny-Gaussian-morphology framework. J. Navig.
**2021**, 74, 1252–1266. [Google Scholar] [CrossRef] - Yang, D.; Wu, L.; Wang, S.; Jia, H.; Li, K.X. How big data enriches maritime research–a critical review of automatic identification system (AIS) data applications. Transp. Rev.
**2019**, 39, 755–773. [Google Scholar] [CrossRef] - Yan, H.; Cui, Z.; Chen, X.; Ma, X. Distributed Multi-Agent Deep Reinforcement Learning for Multi-Line Dynamic Bus Timetable Optimization; IEEE: Manhattan, NY, USA, 2022; p. 1. [Google Scholar] [CrossRef]
- Kurekin, A.A.; Loveday, B.R.; Clements, O.; Quartly, G.D.; Miller, P.I.; Wiafe, G.; Adu Agyekum, K. Operational monitoring of illegal fishing in Ghana through exploitation of satellite earth observation and AIS data. Remote Sens.
**2019**, 11, 293. [Google Scholar] [CrossRef] - Xiao, G.; Wang, T.; Chen, X.; Zhou, L. Evaluation of Ship Pollutant Emissions in the Ports of Los Angeles and Long Beach. J. Mar. Sci. Eng.
**2022**, 10, 1206. [Google Scholar] [CrossRef] - Doğan, E. LSTM training set analysis and clustering model development for short-term traffic flow prediction. Neural Comput. Appl.
**2021**, 33, 11175–11188. [Google Scholar] [CrossRef] - Chen, X.; Wu, S.; Shi, C.; Huang, Y.; Yang, Y.; Ke, R.; Zhao, J. Sensing Data Supported Traffic Flow Prediction via Denoising Schemes and ANN: A Comparison. IEEE Sens. J.
**2020**, 20, 14317–14328. [Google Scholar] [CrossRef] - Zhou, Y.; Daamen, W.; Vellinga, T.; Hoogendoorn, S.P. Ship classification based on ship behavior clustering from AIS data. Ocean Eng.
**2019**, 175, 176–187. [Google Scholar] [CrossRef] - Bai, X.; Cheng, L.; Iris, Ç. Data-driven financial and operational risk management: Empirical evidence from the global tramp shipping industry. Transp. Res. Part E Logist. Transp. Rev.
**2022**, 158, 102617. [Google Scholar] [CrossRef] - Duan, H.; Ma, F.; Miao, L.; Zhang, C. A semi-supervised deep learning approach for vessel trajectory classification based on AIS data. Ocean Coast. Manag.
**2022**, 218, 106015. [Google Scholar] [CrossRef] - Venturini, G.; Iris, Ç.; Kontovas, C.A.; Larsen, A. The multi-port berth allocation problem with speed optimization and emission considerations. Transp. Res. Part D Transp. Environ.
**2017**, 54, 142–159. [Google Scholar] [CrossRef] - Toscano, D.; Murena, F.; Quaranta, F.; Mocerino, L. Assessment of the impact of ship emissions on air quality based on a complete annual emission inventory using AIS data for the port of Naples. Ocean Eng.
**2021**, 232, 109166. [Google Scholar] [CrossRef] - Chen, X.; Ling, J.; Yang, Y.; Zheng, H.; Xiong, P.; Postolache, O.; Xiong, Y. Ship Trajectory Reconstruction from AIS Sensory Data via Data Quality Control and Prediction. Math. Probl. Eng.
**2020**, 2020, 7191296. [Google Scholar] [CrossRef] - Xueqing, Z.; Zhansong, Z.; Chaomo, Z. Bi-LSTM Deep Neural Network Reservoir Classification Model Based on the Innovative Input of Logging Curve Response Sequences. IEEE Access
**2021**, 9, 19902–19915. [Google Scholar] [CrossRef] - Kashinath, S.A.; Mostafa, S.A.; Mustapha, A.; Mahdin, H.; Lim, D.; Mahmoud, M.A.; Mohammed, M.A.; Al-Rimy, B.A.S.; Fudzee, M.F.M.; Yang, T.J. Review of data fusion methods for real-time and multi-sensor traffic flow analysis. IEEE Access
**2021**, 9, 51258–51276. [Google Scholar] [CrossRef] - Wu, Y.; Tan, H.; Qin, L.; Ran, B.; Jiang, Z. A hybrid deep learning based traffic flow prediction method and its understanding. Transp. Res. Part C Emerg. Technol.
**2018**, 90, 166–180. [Google Scholar] [CrossRef] - Chen, C.; Liu, Z.; Wan, S.; Luan, J.; Pei, Q. Traffic Flow Prediction Based on Deep Learning in Internet of Vehicles. IEEE Trans. Intell. Transp. Syst.
**2021**, 22, 3776–3789. [Google Scholar] [CrossRef] - Alahi, A.; Goel, K.; Ramanathan, V.; Robicquet, A.; Fei-Fei, L.; Savarese, S. Social lstm: Human trajectory prediction in crowded spaces. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Las Vegas, NV, USA, 27–30 June 2016; pp. 961–971. [Google Scholar]
- Gupta, A.; Johnson, J.; Fei-Fei, L.; Savarese, S.; Alahi, A. Social gan: Socially acceptable trajectories with generative adversarial networks. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Salt Lake City, UT, USA, 18–23 June 2018; pp. 2255–2264. [Google Scholar]

hardware | CPU | Intel(R) Core (TM) i7–8750H |

frequency | 2.2 GHz | |

RAM | 16 G | |

GPU | NVIDA GeForce GTX 1050 Ti | |

software | OS | Windows 10 |

language | Python 3.6 | |

deep learning framework | tensoflow1.14/Keras 2.2.5 | |

CUDA version | 10.1 |

Epoch | Learning Rate | Batch Size |
---|---|---|

2000 | 1.00 × 10^{−4} | 64 |

Model | Input Feature Number | Hidden Layer Node | Ship Number |
---|---|---|---|

Bi-LSTM | 24 | 200 | single-ship |

8 | 200 | multiple-ship | |

LSTM | 48 | 120 | single-ship |

24 | 120 | multiple-ship | |

SOC-LSTM | 48 | 128 | single-ship |

24 | 128 | multiple-ship | |

SOC-GAN | 48 | 64 | single-ship |

24 | 64 | multiple-ship |

Ground Truth | Bi-LSTM | LSTM | SOC-LSTM | SOC-GAN |
---|---|---|---|---|

32.00222 | 32.00221 | 32.00219 | 32.00168 | 32.00216 |

32.00220 | 32.00219 | 32.00217 | 32.00162 | 32.00200 |

32.00218 | 32.00217 | 32.00215 | 32.00154 | 32.00175 |

32.00216 | 32.00214 | 32.00213 | 32.00126 | 32.00155 |

32.00214 | 32.00213 | 32.00211 | 32.00117 | 32.00153 |

32.00211 | 32.00210 | 32.00209 | 32.00101 | 32.00153 |

32.00208 | 32.00207 | 32.00207 | 32.00076 | 32.00124 |

Ground Truth | Bi-LSTM | LSTM | SOC-LSTM | SOC-GAN |
---|---|---|---|---|

120.66696 | 120.66694 | 120.66698 | 120.66661 | 120.66724 |

120.66682 | 120.66679 | 120.66684 | 120.66634 | 120.66722 |

120.66668 | 120.66665 | 120.66669 | 120.66605 | 120.66687 |

120.66654 | 120.66649 | 120.66656 | 120.6657 | 120.66670 |

120.66639 | 120.66637 | 120.66642 | 120.66528 | 120.66689 |

120.66625 | 120.66622 | 120.66629 | 120.66480 | 120.66666 |

120.66611 | 120.66606 | 120.66618 | 120.66456 | 120.66652 |

MAE | MAPE | MSE | |
---|---|---|---|

Bi-LSTM | 4.66 × 10^{−5} | 6.74 × 10^{−5} | 1.43 × 10^{−9} |

LSTM | 5.76 × 10^{−5} | 1.15 × 10^{−4} | 1.94 × 10^{−9} |

SOC-LSTM | 4.46 × 10^{−4} | 8.66 × 10^{−4} | 4.63 × 10^{−7} |

SOC-GAN | 1.64 × 10^{−1} | 4.13 × 10^{−4} | 7.79 × 10^{−2} |

MAE | MAPE | MSE | |
---|---|---|---|

Bi-LSTM | 6.46 × 10^{−5} | 9.07 × 10^{−5} | 3.98 × 10^{−9} |

LSTM | 1.87 × 10^{−4} | 1.31 × 10^{−4} | 1.02 × 10^{−8} |

SOC-LSTM | 3.57 × 10^{−4} | 7.98 × 10^{−4} | 3.31 × 10^{−7} |

SOC-GAN | 1.70 × 10^{−4} | 3.40 × 10^{−4} | 7.44 × 10^{−8} |

Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. |

© 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).

## Share and Cite

**MDPI and ACS Style**

Chen, X.; Wei, C.; Zhou, G.; Wu, H.; Wang, Z.; Biancardo, S.A.
Automatic Identification System (AIS) Data Supported Ship Trajectory Prediction and Analysis via a Deep Learning Model. *J. Mar. Sci. Eng.* **2022**, *10*, 1314.
https://doi.org/10.3390/jmse10091314

**AMA Style**

Chen X, Wei C, Zhou G, Wu H, Wang Z, Biancardo SA.
Automatic Identification System (AIS) Data Supported Ship Trajectory Prediction and Analysis via a Deep Learning Model. *Journal of Marine Science and Engineering*. 2022; 10(9):1314.
https://doi.org/10.3390/jmse10091314

**Chicago/Turabian Style**

Chen, Xinqiang, Chenxin Wei, Guiliang Zhou, Huafeng Wu, Zhongyu Wang, and Salvatore Antonio Biancardo.
2022. "Automatic Identification System (AIS) Data Supported Ship Trajectory Prediction and Analysis via a Deep Learning Model" *Journal of Marine Science and Engineering* 10, no. 9: 1314.
https://doi.org/10.3390/jmse10091314