arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2601.14880eess.SYcs.SY
收藏

安全关键性自动驾驶的应急规划:综述与展望

Contingency Planning for Safety-Critical Autonomous Vehicles: A Review and Perspectives

Carnegie Mellon University(卡内基梅隆大学);The Hong Kong University of Science and Technology(香港科技大学);Delft University of Technology(代尔夫特理工大学)

Lei Zheng, Luyao Zhang, Peiqi Yu, Yifan Sun, Sergio Grammatico, Jun Ma, Changliu Liu

AI总结:

本文综述了安全关键性自动驾驶的应急规划,提出统一的混合控制框架,区分反应安全与主动安全范式,并指出内部故障与外部不确定性处理方法的差异。

摘要:

应急规划是使自动驾驶车辆(AVs)能够预见并缓解离散、高影响的危险事件(如传感器故障和对抗性交互)的架构能力。本文对这一领域进行了全面的综述,将碎片化的文献整合成一个统一的逻辑-条件混合控制框架。在此形式化体系中,我们将方法分为两种不同的范式:反应安全,即通过执行安全约束或执行故障安全操作来应对已发生的危险;以及主动安全,即通过分支潜在的模式转换来优化未来的选择。此外,我们提出了一种细粒度的分类法,将领域分为外部应急(环境和交互危险)和内部应急(系统故障)。通过批判性的比较分析,我们揭示了根本性的结构分歧:内部故障主要通过反应性故障安全机制处理,而外部交互不确定性越来越多地需要主动分支策略。此外,我们识别出一种关键的方法论分歧:物理危险通常通过形式化保证来管理,而语义和分布外异常目前严重依赖经验验证。我们最后指出在弥合理论保证与实际验证之间差距的开放挑战,倡导混合架构和标准化基准测试,将应急规划从制定过渡到可验证的现实部署。

英文摘要:

Contingency planning is an architectural capability that enables autonomous vehicles (AVs) to anticipate and mitigate safety risks arising from safety-relevant logical-mode transitions, such as sensor outages and uncertain interactions. This paper presents a comprehensive survey of the field, synthesizing fragmented literature into a unified logic-conditioned hybrid control framework. Within this formalism, we categorize approaches into two distinct paradigms: Reactive Safety enforces safety constraints or activates fallback actions after a contingency mode has been observed or reliably inferred, whereas Proactive Safety preserves recourse across unresolved mode alternatives through non-anticipative branching. In addition, we propose a fine-grained taxonomy that partitions the landscape into external contingencies (environmental and interactive hazards) and internal contingencies (system faults). Our analysis reveals a fundamental structural divergence: internal faults are predominantly addressed via reactive fail-safe mechanisms, whereas external interaction uncertainties increasingly require proactive branching strategies. Furthermore, we identify a critical methodological divergence: whereas physical hazards are typically managed with formal guarantees, semantic and out-of-distribution anomalies currently rely heavily on empirical validation. We conclude by identifying the open challenges in bridging the gap between theoretical guarantees and practical validation, advocating for hybrid architectures and standardized benchmarking to transition contingency planning from formulation to certifiable real-world deployment.

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。

↑