SS 06
The practical deployment of Machine Learning (ML) and Artificial Intelligence (AI) for Prognostics and Health Management (PHM) of aerospace and avionic systems is often hindered by “black-box” limitations, including data scarcity in extreme operating scenarios, physical inconsistency with aerospace engineering principles, the lack of robust explainability, and poor generalizability to unseen flight and mission conditions. As a critical branch of interpretable intelligent frameworks, Physics-Informed Machine Learning (PIML) and Explainable AI (XAI) integrate prior physics-based knowledge, failure mechanisms, domain expertise, and airworthiness requirements directly into the PHM learning process, breaking the bottlenecks of traditional data-driven models for safety-critical aerospace equipment.
This special session explores how XAI and PIML enable aerospace and avionic PHM to move toward physically grounded, interpretable, certifiable and trustworthy AI solutions, bridging the gap between theoretical algorithm innovation and industrial engineering deployment. We invite global researchers, industry practitioners and certification experts to share recent advancements in interpretable ML/PIML methodologies, as well as their field-specific applications to the reliability, health assessment and predictive maintenance of complex aerospace and avionic systems.
在航空航天与航空电子系统故障预测与健康管理（PHM）的实际部署中，机器学习与人工智能的"黑箱"特性常面临多重制约：极端运行场景下的数据稀缺性、与航空航天工程原理的物理一致性缺失、可解释性不足，以及面对未知飞行任务条件时的泛化能力薄弱。作为可解释智能框架的重要分支，物理信息机器学习和可解释人工智能通过将基于物理的先验知识、失效机理、领域专业知识和适航要求直接嵌入PHM学习过程，突破了传统数据驱动模型在安全关键型航空航天装备应用中的瓶颈。
本专题旨在探讨XAI与PIML如何推动航空航天PHM向物理可循、可解释、可认证且可信的AI解决方案演进，从而弥合理论算法创新与工业工程应用之间的鸿沟。我们诚邀全球研究人员、行业实践者及认证专家，分享在可解释机器学习/PIML方法论领域的最新突破，及其在复杂航空航天与航空电子系统可靠性评估、健康状态监测与预测性维护等具体场景中的创新应用。
征稿主题包括但不限于 Topics of interest include, but are not limited to：
- - 物理信息驱动的PHM模型与算法 (Physics-Informed/Hybrid Models for PHM)
- - 航空航天PHM可解释AI算法及框架 (Explainable AI Methods & Frameworks in aviation and aerosoace PHM )
- - 面向极端场景与小样本的物理信息PHM方法 Physics-Informed PHM Methods for Extreme Scenarios and Small Data
- - 稀缺数据与仿真增强方法 (Data Augmentation & Simulation-Based Learning)
- - 多模态传感与知识融合 (Multi-Modal Sensor Fusion & Knowledge Integration)
- - 数字孪生与仿真辅助PHM (Digital Twin & Simulation-Enriched PHM)
- - 融合失效机理与领域知识的可解释PHM建模
- - 可解释AI驱动的预测性维护与维修决策 Explainable AI-Driven Predictive Maintenance and Maintenance Decision-Making
- - XAI与PIML在典型航空航天装备中的应用验证 Application and Validation of XAI and PIML in Typical Aerospace Equipment
