报告名称： Aircraft Virtual Flight Testing and Certification in Off-nominal Multi-factorial Situations（非典型条件多因素影响下的飞行器虚拟试飞及验证）

 主讲人：伊万·伯顿教授（法国AIXTREE SAS）

 主办单位：航空学院、飞行器先进设计技术国防重点学科实验室、校科协、国合处

 报告时间：2025年10月24日 上午9:00 -11：40，14:00-18:00

 报告内容简介：

 第一天的课程主要探讨如何构建非典型多重因素飞行领域的分析建模，该领域以复杂场景为特征，例如连锁反应事故和"维度灾难"。课程讨论了经典飞行研究技术的局限性，并提出了一个增强型的飞行测试系统，其核心原则是：通过虚拟、自主、快速地对"替代未来"进行探索，实现极端测试案例的"去硬件化"。课程详细阐述了该领域的两级知识模型，定义了飞行事件、过程和情景等要素，并引入了多重因素风险假设。一个核心概念是情境树，它模拟了飞行员长时记忆和专业能力发展的分支结构，旨在实现一种理想的分形增长模型。课程进一步介绍了使用安全色标、部分/整体安全谱、飞行安全指数和安全拓扑等工具进行飞行安全度量和绘图的方法。通过对系统动力学模型的概述，解释了其组成部分，包括飞行物理学、飞行员与自动化模型及其显著特点。课程还讨论了软件实现、开发历史、算法和技术规范，以及在不同地区和飞机类型上的过往应用案例，并通过对起飞、着陆等不同飞行阶段的验证示例来评估模型的有效性。

 This course explores the formulation of research tasks and the analysis of the off-nominal multifactorial flight domain, characterized by complex scenarios like chain-reaction accidents and the curse of dimensionality. It critiques the limitations of classic flight research techniques and proposes an augmented Flight Test Cycle (FTC) centered on a key principle: the 'de-materialization' of extreme corner cases through virtual, autonomous, fast-time exploration of alternative futures. The curriculum details a two-level knowledge model of this domain, defining elements like flight events, processes, and situations, and introduces multifactorial risk hypotheses. A core concept is the situational (tactical) tree, which models the branching structure of a pilot's long-term memory and expertise development, aiming for an ideal fractal growth model. The course further covers the measurement and mapping of flight safety using tools like safety color coding, partial/integral safety spectra, a flight safety index, and safety topology. An overview of the System Dynamics model explains its components, including flight physics, human pilot and automation models, and its distinguishing features. The software implementation, development history, algorithms, and technical specifications are discussed, along with past applications across various geographies and aircraft types, supported by validation examples from different flight phases like takeoff and landing, assessing the model's validity.

 报告时间：2025年10月25日 上午9:00 -11：40，14:00-18:00

 报告内容简介：

 本部分详细阐述了用于早期探索复杂飞行领域的虚拟自主快速飞行测试周期的具体实施，概述了其整体布局、核心处理代理、数据流和标准化算法。它明确了对飞机全面“参数定义”的要求，包括其组成部分、数据来源、气动数据的“拼接”、质量指标和自动生成，以及其在系统模型中的应用。课程描述了基于适航规章等来源为不同飞行阶段设计基线飞行场景的过程。这一过程涉及对风险因素进行系统化分类，并为不同概念机构型在各种飞行状态下的多因素风险假说进行规划。工作流程进一步扩展到规划和运行快速模拟实验、控制误差并管理情境树的增长，最后是知识挖掘和映射，即生成输出数据格式库并构建特定的知识图谱以解读结果。

 This segment details the implementation of a virtual autonomous fast-time flight test cycle for the early exploration of complex flight domains, outlining its general layout, key processing agents, data flows, and standardized algorithm. It specifies the requirements for a comprehensive aircraft 'parametric definition,' covering its components, data sourcing, the 'stitching' of aerodynamic data, quality metrics, and automated generation, along with its utilization within the System model. The process for designing baseline flight scenarios from sources like regulations and manuals is described for various flight phases. It involves the formalization of a taxonomy of risk factors and the planning of multifactorial risk hypotheses for notional aircraft types across different flight regimes. The workflow extends to planning and running fast-time simulation experiments, controlling for errors and managing the situational tree's growth, and concludes with knowledge mining and mapping, which involves generating a library of output data formats and constructing specific knowledge maps to interpret the results.

 报告时间：2025年10月26日 上午9:00 -11：40，14:00-18:00

 报告内容简介：

 课程的最终阶段从理论过渡到实践分析，涵盖了使用概念机构型进行的单情境和多情境分析。这部分内容包括将风险因素和多因素假说应用于一系列基线场景——包括各种起飞、着陆、平飞和特殊机动——以评估良性及复杂案例。随后，课程展望了该技术的未来发展和应用前景，例如先进的陆基和机载系统、利用情境树和整体安全谱进行运行领域筛查，以及用于预测和预防灾难性事故的关键概念——动态安全窗口。课程最后评估了VFTC技术的显著特点、用户收益，以及其内在的研究挑战、潜在缺陷和局限性，并进行了总结和学习评估。

 This final phase of the course transitions from theory to practical analysis, covering both single and multiple situation analyses using notional aircraft types. It involves applying risk factors and multifactorial hypotheses to a range of baseline scenarios—including various take-off, landing, level flight, and special maneuvers—to assess both benign and complex cases. The curriculum then looks forward to future developments and prospective applications of the technology, such as advanced ground and onboard systems, the use of situational trees and integral safety spectra for domain screening, and the critical concept of dynamic safety windows for predicting and preventing catastrophic accidents. The course concludes by evaluating the distinguishing features, user benefits, and the inherent research challenges, pitfalls, and limitations of the VFTC technology, followed by a summary and learning assessment.

 报告人简介：

伊万·伯顿教授拥有四十年跨文化研究与学术经验，其职业生涯遍布多国顶尖机构：包括美国佐治亚理工学院航空航天工程学院、英国克兰菲尔德大学航空学院、苏联拉脱维亚里加民用航空工程学院空气动力学与飞行动力学系、俄罗斯西伯利亚航空研究所飞机空气动力学与飞行动力学研究部、俄罗斯新西伯利亚国立技术大学航空航天工程学院、法国AIXTREE SAS等。他的专业能力涵盖高保真数学建模、自主快速飞行仿真、拟人化人工智能，以及通过知识挖掘与图谱构建来预测"人类操作员-自动化系统-飞行器-运行环境"这一系统在多重风险因素构成的复杂未知情境中的动态行为与安全性能。这些技术已应用于34种飞行器型号与设计项目，涵盖固定翼与旋翼机、倾转旋翼机，以及亚音速、超音速和高超音速飞行器。布尔登博士当前的研究与学术重点聚焦于超强效且成本可控的拟人化人工智能技术，用于飞行安全预测与防护、系统行为异常识别、有人/无人驾驶航空器及集群系统中飞行员与人工智能的认知接口原型设计及联合决策机制开发。

Dr. Ivan BURDUN has 40 years of cross-cultural research and academic experience at the School of Aerospace Engineering at the Georgia Institute of Technology (USA), the College of Aeronautics at the Cranfield University (UK), the Department of Aerodynamics and Flight Dynamics at the Riga Civil Aviation Engineering Institute (USSR, Latvia), the Aircraft Aerodynamics and Flight Dynamics Research Division of the Siberian Aeronautical Research Institute (Russia), the School of Aerospace Engineering of the Novosibirsk State Technical University (Russia), AIXTREE SAS (France), and other organisations. His competences include high-fidelity mathematical modeling, autonomous fast-time flight simulation, anthropomorphic artificial intelligence, knowledge mining and mapping for predicting the 'human operator - automaton - aircraft - operating environment' system dynamics and safety performance in complex and unknown situations with multiple risk factors. These techniques have been applied to 34 aircraft types and design projects: fixed- and rotary-wing, tilt-rotor; sub-, super- and hypersonic. Dr. BURDUN's current research and academic focus is on super-competent (anthropomorphic) affordable AI techniques for flight safety prediction and protection, identification of the anomalies in the system behavior, prototyping of pilot-AI cognitive interface and joint decision-making for manned and unmanned vehicles and swarms.