Research

Paper

AI LLM March 03, 2026

ShipTraj-R1: Reinforcing Ship Trajectory Prediction in Large Language Models via Group Relative Policy Optimization

Authors

Yang Zhan, Yunhao Li, Zhang Chao, Yuxu Lu, Yan Li

Abstract

Recent advancements in reinforcement fine-tuning have significantly improved the reasoning ability of large language models (LLMs). In particular, methods such as group relative policy optimization (GRPO) have demonstrated strong capabilities across various fields. However, applying LLMs to ship trajectory prediction remains largely unexplored. In this paper, we propose ShipTraj-R1, a novel LLM-based framework that reformulates ship trajectory prediction as a text-to-text generation problem. (1) We design a dynamic prompt containing trajectory information about conflicting ships to guide the model to achieve adaptive chain-of-thought (CoT) reasoning. (2) We introduce a comprehensive rule-based reward mechanism to incentivize the reasoning format and prediction accuracy of the model. (3) Our ShipTraj-R1 is reinforced through the GRPO mechanism guided by domain-specific prompts and rewards, and utilizes the Qwen3 as the model backbone. Extensive experimental results on two complex and real-world maritime datasets show that the proposed ShipTraj-R1 achieves the least error compared with state-of-the-art deep learning and LLM-based baselines.

Metadata

arXiv ID: 2603.02939
Provider: ARXIV
Primary Category: cs.AI
Published: 2026-03-03
Fetched: 2026-03-04 03:41

Related papers

Raw Data (Debug)
{
  "raw_xml": "<entry>\n    <id>http://arxiv.org/abs/2603.02939v1</id>\n    <title>ShipTraj-R1: Reinforcing Ship Trajectory Prediction in Large Language Models via Group Relative Policy Optimization</title>\n    <updated>2026-03-03T12:48:40Z</updated>\n    <link href='https://arxiv.org/abs/2603.02939v1' rel='alternate' type='text/html'/>\n    <link href='https://arxiv.org/pdf/2603.02939v1' rel='related' title='pdf' type='application/pdf'/>\n    <summary>Recent advancements in reinforcement fine-tuning have significantly improved the reasoning ability of large language models (LLMs). In particular, methods such as group relative policy optimization (GRPO) have demonstrated strong capabilities across various fields. However, applying LLMs to ship trajectory prediction remains largely unexplored. In this paper, we propose ShipTraj-R1, a novel LLM-based framework that reformulates ship trajectory prediction as a text-to-text generation problem. (1) We design a dynamic prompt containing trajectory information about conflicting ships to guide the model to achieve adaptive chain-of-thought (CoT) reasoning. (2) We introduce a comprehensive rule-based reward mechanism to incentivize the reasoning format and prediction accuracy of the model. (3) Our ShipTraj-R1 is reinforced through the GRPO mechanism guided by domain-specific prompts and rewards, and utilizes the Qwen3 as the model backbone. Extensive experimental results on two complex and real-world maritime datasets show that the proposed ShipTraj-R1 achieves the least error compared with state-of-the-art deep learning and LLM-based baselines.</summary>\n    <category scheme='http://arxiv.org/schemas/atom' term='cs.AI'/>\n    <published>2026-03-03T12:48:40Z</published>\n    <arxiv:comment>Accepted by the 30th Pacific-Asia Conference on Knowledge Discovery and Data Mining (PAKDD2026)</arxiv:comment>\n    <arxiv:primary_category term='cs.AI'/>\n    <author>\n      <name>Yang Zhan</name>\n    </author>\n    <author>\n      <name>Yunhao Li</name>\n    </author>\n    <author>\n      <name>Zhang Chao</name>\n    </author>\n    <author>\n      <name>Yuxu Lu</name>\n    </author>\n    <author>\n      <name>Yan Li</name>\n    </author>\n  </entry>"
}