Intent prediction of vessels in intersection waterway based on learning vessel motion patterns with early observations
Ocean Engineering, ISSN: 0029-8018, Vol: 232, Page: 109154
2021
- 27Citations
- 26Captures
Metric Options: Counts1 Year3 YearSelecting the 1-year or 3-year option will change the metrics count to percentiles, illustrating how an article or review compares to other articles or reviews within the selected time period in the same journal. Selecting the 1-year option compares the metrics against other articles/reviews that were also published in the same calendar year. Selecting the 3-year option compares the metrics against other articles/reviews that were also published in the same calendar year plus the two years prior.
Example: if you select the 1-year option for an article published in 2019 and a metric category shows 90%, that means that the article or review is performing better than 90% of the other articles/reviews published in that journal in 2019. If you select the 3-year option for the same article published in 2019 and the metric category shows 90%, that means that the article or review is performing better than 90% of the other articles/reviews published in that journal in 2019, 2018 and 2017.
Citation Benchmarking is provided by Scopus and SciVal and is different from the metrics context provided by PlumX Metrics.
Example: if you select the 1-year option for an article published in 2019 and a metric category shows 90%, that means that the article or review is performing better than 90% of the other articles/reviews published in that journal in 2019. If you select the 3-year option for the same article published in 2019 and the metric category shows 90%, that means that the article or review is performing better than 90% of the other articles/reviews published in that journal in 2019, 2018 and 2017.
Citation Benchmarking is provided by Scopus and SciVal and is different from the metrics context provided by PlumX Metrics.
Article Description
The situational awareness in intersection waterways should be improved given the numerous accidents occurring in these areas. We propose a deep learning model for predicting the sailing intent in the intersection waterway because ship–ship collisions in such areas mainly occur by incorrectly interpreting the intents of other vessels. We model the continuous motions of vessels before entering an intersection and estimate the most likely intent of the vessel given early observations. From a statistical analysis of observed motions, we found that vessel motion tends to be highly related to past trajectories over long periods. Therefore, we introduce a recurrent neural network architecture, the accumulated long short-term memory (ALSTM), which overcomes the limitations of the conventional LSTM by adopting skip connections and an adaptive memory module. Using these two extensions, the present memory cell interacts with historical data to store and use relevant memory details even after long periods. Furthermore, the strong generalization ability of the proposed accumulated LSTM enables a high-level representation of the uncertain and diverse motions of individual vessels, thus contributing to relate various ship motion patterns and intents. For validation, the proposed model is trained and tested in a naturalistic data set with exemplary results. In addition, we compare the proposed accumulated LSTM with the conventional LSTM and a hidden Markov model. Experimental results demonstrate the improved performance of the proposed accumulated LSTM in terms of accuracy and real-time response.
Bibliographic Details
http://www.sciencedirect.com/science/article/pii/S0029801821005874; http://dx.doi.org/10.1016/j.oceaneng.2021.109154; http://www.scopus.com/inward/record.url?partnerID=HzOxMe3b&scp=85107440782&origin=inward; https://linkinghub.elsevier.com/retrieve/pii/S0029801821005874; https://dx.doi.org/10.1016/j.oceaneng.2021.109154
Elsevier BV
Provide Feedback
Have ideas for a new metric? Would you like to see something else here?Let us know