Maritime Target Trajectory Prediction Model
Based on the RNN Network
Jialong Jin, Wei Zhou
(&)
, and Baichen Jiang
Naval Aviation University, Shandong 264000, China
jjldwb@sina.cn, yeaweam@163.com, 704011136@qq.com
Abstract. The number of ships has increased in recent years and the requirements for maritime security protection have become more stringent. It is
important to improve the prediction accuracy and efficiency of maritime target in
order to sustain maritime security. According to the real-time, efficiency and
accuracy requirements of maritime target trajectory prediction, a prediction
model based on RNN network is proposed to realize maritime target trajectory
prediction based on AIS data. This paper uses AIS data between Longkou and
Dalian to conduct experiments, and compares the prediction effects of two
network models RNN and LSTM on the maritime target trajectory. It proves that
the RNN network model has higher prediction accuracy and stronger learning
ability. Based on the experimental results, this paper analyzes the characteristics
of deep learning in long-term prediction and historical data dependence at the
same time.
Keywords: Trajectory prediction Á Deep learning Á Automatic identification
system (AIS) Á RNN Á LSTM
1 Introduction
With the increasingly complex maritime situation, maritime security needs are
increasing. As an important means to ensure maritime security, the trajectory prediction
of maritime targets has been widely studied. At present, a lot of research has been done
on target trajectory prediction, Gambs [1] proposed an extended Mobility Markov
Chain in order to predict next place based on the n previous locations visited. Its
accuracy can reach from 70 to 95%. But it cost a lot in terms of computation and space.
Song [2] finds that a 93% potential predictability to predict the spread of human with
measuring the entropy of each individual’s trajectory. Based on the findings, Qiao [3]
proposed to predict target’s trajectory with Hidden Markov Model. But he does not
consider the problem of efficiency under the background of Big Data. Then, he proposed a new dynamic trajectory prediction algorithm [4] based on Kalman filter [5],
which can guarantee time performance. Besides, there are also many deep learning
algorithms [6–8], but most of them focus on visual prediction. Quan [9] used LSTM
network to predict trajectory based on AIS data and proved that it is better than BP
network. This paper proposed a simplified RNN network, which is simpler than LSTM
and behaves better.
© Springer Nature Singapore Pte Ltd. 2020
Q. Liang et al. (Eds.): Artificial Intelligence in China, LNEE 572, pp. 334–342, 2020.
https://doi.org/10.1007/978-981-15-0187-6_39
Based on the RNN Network
Jialong Jin, Wei Zhou
(&)
, and Baichen Jiang
Naval Aviation University, Shandong 264000, China
jjldwb@sina.cn, yeaweam@163.com, 704011136@qq.com
Abstract. The number of ships has increased in recent years and the requirements for maritime security protection have become more stringent. It is
important to improve the prediction accuracy and efficiency of maritime target in
order to sustain maritime security. According to the real-time, efficiency and
accuracy requirements of maritime target trajectory prediction, a prediction
model based on RNN network is proposed to realize maritime target trajectory
prediction based on AIS data. This paper uses AIS data between Longkou and
Dalian to conduct experiments, and compares the prediction effects of two
network models RNN and LSTM on the maritime target trajectory. It proves that
the RNN network model has higher prediction accuracy and stronger learning
ability. Based on the experimental results, this paper analyzes the characteristics
of deep learning in long-term prediction and historical data dependence at the
same time.
Keywords: Trajectory prediction Á Deep learning Á Automatic identification
system (AIS) Á RNN Á LSTM
1 Introduction
With the increasingly complex maritime situation, maritime security needs are
increasing. As an important means to ensure maritime security, the trajectory prediction
of maritime targets has been widely studied. At present, a lot of research has been done
on target trajectory prediction, Gambs [1] proposed an extended Mobility Markov
Chain in order to predict next place based on the n previous locations visited. Its
accuracy can reach from 70 to 95%. But it cost a lot in terms of computation and space.
Song [2] finds that a 93% potential predictability to predict the spread of human with
measuring the entropy of each individual’s trajectory. Based on the findings, Qiao [3]
proposed to predict target’s trajectory with Hidden Markov Model. But he does not
consider the problem of efficiency under the background of Big Data. Then, he proposed a new dynamic trajectory prediction algorithm [4] based on Kalman filter [5],
which can guarantee time performance. Besides, there are also many deep learning
algorithms [6–8], but most of them focus on visual prediction. Quan [9] used LSTM
network to predict trajectory based on AIS data and proved that it is better than BP
network. This paper proposed a simplified RNN network, which is simpler than LSTM
and behaves better.
© Springer Nature Singapore Pte Ltd. 2020
Q. Liang et al. (Eds.): Artificial Intelligence in China, LNEE 572, pp. 334–342, 2020.
https://doi.org/10.1007/978-981-15-0187-6_39