改进航空安全分析：使用带 Group Relative Policy Optimization 的强化学习进行自动化 HFACS 分类
Improving Aviation Safety Analysis: Automated HFACS Classification Using Reinforcement Learning with Group Relative Policy Optimization
AI总结：
提出基于 GRPO 强化学习微调 Llama-3.1 8B 的自动化 HFACS 分类框架，结合多组件奖励与合成数据，显著提升准确率并超过 GPT-5-mini 等大模型。
摘要：
分析航空事故背后的人为因素对于预防未来事件至关重要，但使用 Human Factors Analysis and Classification System（HFACS，人为因素分析与分类系统）的传统方法受到可扩展性和一致性的限制。为解决这一问题，我们提出一种用于航空安全分析的自动化 HFACS 分类框架，该框架利用带 Group Relative Policy Optimization（GRPO，组相对策略优化）的强化学习来微调 Llama-3.1 8B 语言模型。我们的方法包含为航空安全分析定制的多组件奖励系统，并整合合成数据生成以克服事故数据集中的类别不平衡。由此得到的 GRPO 优化模型取得了显著性能提升，包括 exact match accuracy（精确匹配准确率）提升 350%（从 0.0400 增至 0.1800），以及 partial match accuracy（部分匹配准确率）提升至 0.8800。值得注意的是，我们的专用模型在关键指标上优于包括 GPT-5-mini 和 Gemini-2.5-fiash 在内的最先进 LLM（Large Language Models，大语言模型）。本研究还提出将多标签 HFACS 分类问题中的精确匹配准确率作为评估语言模型高级推理能力的新基准方法。最终，我们的工作验证了较小的、经领域优化的模型能够为关键安全分析提供计算高效且更优的解决方案。该方法使在资源受限的边缘设备上实现强大、低延迟的部署成为可能。
英文摘要：
Analyzing the human factors behind aviation accidents is crucial for preventing future incidents, yet traditional methods using the Human Factors Analysis and Classification System (HFACS) are limited by scalability and consistency. To address this, we introduce an automated HFACS classification framework for aviation safety analysis that utilizes Reinforcement Learning with Group Relative Policy Optimization (GRPO) to fine-tune a Llama-3.1 8B language model. Our approach incorporates a multi-component reward system tailored for aviation safety analysis and integrates synthetic data generation to overcome class imbalance in accident datasets. The resulting GRPO-optimized model achieved noticeable performance gains, including a 350% increase in exact match accuracy (from 0.0400 to 0.1800) and an improved partial match accuracy of 0.8800. Significantly, our specialized model outperforms state-of-the-art LLMs (Large Language Models), including GPT-5-mini and Gemini-2.5-fiash, on key metrics. This research also proposes exact match accuracy in multi-label HFACS classification problem as a new benchmarking methodology to evaluate the advanced reasoning capabilities of language models. Ultimately, our work validates that smaller, domain-optimized models can provide a computationally efficient and better solution for critical safety analysis. This approach makes powerful, low-latency deployment on resource-constrained edge devices feasible.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
