学术信息

【航宇大讲堂】High-Fidelity Mathematical Modelling, Fast-Time Simulation and Artificial Intelligence Techniques for Flight Safety Protection in '9/11' Class Scenarios

发布时间:2020-11-16浏览次数:1472发布者:颜士轩来源:南京航空航天大学

报告题目:High-Fidelity Mathematical Modelling, Fast-Time Simulation and  Artificial Intelligence Techniques for Flight Safety Protection in '9/11' Class Scenarios

报告人:Dr. Ivan Y. BURDUN(AIXTREE S.A.S)

报告时间:2020年11月16日(周一)17:00-18:00

报告地点:腾讯视频会议ID:658 5662 4196

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、国际合作处、科协

报告内容摘要:

9/11 class scenarios relate to abnormal flight situations characterized by inadequate, a priori catastrophe-prone aircraft control tactics (manual, automated, or automatic).

A knowledge-centered solution approach is presented for predicting and protecting aircraft flight safety under complex conditions with multiple or unknown risks - at and beyond operational constraints. The technique is based on the synergy of high-fidelity mathematical modeling of the pilot/automaton - aircraft - operating environment system dynamics, autonomous fast-time computer experimentation with the model  (M&S), artificial intelligence (AI), knowledge mining and mapping (M&M), and a 'bird's eye view' visual analytics. Its key concepts include: situation scenario, risk factor, multifactorial risk hypothesis, safety palette, fuzzy constraint, integral safety spectrum, situational tree, safety window, safety ‘topology’, ‘last chance for recovery’ point, etc.

 A large set of off-nominal flight paths - what-if variations of the baseline situation's flight path - are automatically generated in advance in fast-time AI-driven M&S experiments. For this purpose, AI 'silicon pilot' model and tree growth control algorithms are used to explore combined effects of related risk (and/or remedial) factors. Situational trees can accumulate extra-large synthetic flying experience - the system's 'alternative futures' built around baseline scenarios.

The technique's advantages are summarized,   however,there are many unsettled issues and prerequisites to be addressed in further studies

报告人简介:

Dr. Ivan Y. BURDUN is the president of AIXTREE S.A.S., Aix-en-Provence, France, form 2013 to date.In 1982,he got PhD degree in Flight Operations awarded by the Higher Attestation Commission (VAC) under the Council of  Ministers of the USSR, Moscow, Russia.He was an Associate Professor (part-time) in the Department of Aero- and Hydro-Dynamics (AGD), School of Aerospace Engineering , Novosibirsk State Technical University ,and was Lead Scientist of Aircraft Aerodynamics and Flight Dynamics Research Division (NIO-1), Siberian Aeronautical. He has published 250+ technical reports, presentations, papers, lecture courses and databases in Russian and English.

He has 40+ years of cross-cultural multidisciplinary research and academic experience in flight  modeling, simulation (M&S) and artificial intelligence (AI) for safety in complex situations.