Agent-based safety risk analysis of future Air Traffic Management designs

报告题目：Agent-based safety risk analysis of future Air Traffic Management designs
主讲人：代尔夫特理工大学教授 / 荷兰国家航空航天实验室首席科学家Henk Blom
报告时间：2018年6月4日-11日；4:30-17:00
报告地点：北京航空航天大学 为民楼320会议室
主讲人简介：
Blom教授于2011年起担任代尔夫特理工大学教授以及空中交通管理安全系主任，并于2004年起担任荷兰国家航天实验室首席科学家。Blom教授在利用随机建模分析理论进行安全风险分析、多传感器数据融合和空中交通管理安全性分析方面有着近30年的经验。他是科学创新发展的领导人，创建了众多创新方法，如交互式多模型滤波，欧洲航空安全组织的贝叶斯多传感器多目标跟踪系统ARTAS（ATM雷达跟踪服务器），基于主体的安全风险分析方法TOPAZ（交通组织摄动法）等。
报告具体日程安排：
1.6月4日（星期一）: Motivation and terminology
Lecture 1: Providing motivation of the relevance of proper safety risk modelling and analysis of operations in Air Traffic. To illustrate the power of agent-based safety risk modelling and analysis in learning an advanced operation like free flight.
Lecture 2: Providing insight in safety terminology, aviation accident statistics, safety regulation and safety management. Hands-on exercise: Apply the material from lectures 1 and 2 to a simple mode of transportation.  
2.6月5日（星期二）: Safety risk assessment basics
Lecture 3: Learning the basic safety risk assessment steps. Objective of a safety risk assessment. Identify the operation considered. Pushing the boundary between imaginable and unimaginable hazards. Hands-on exercise: Evaluate a given safety case against what you learned.
Lecture 4: Construction of scenarios.  Types of safety risk models. Comparison to safety risk criteria.  Hands-on exercise: Continue evaluation of a given safety case against what you learned.
3.6月6日（星期三）: Agent-based modelling semantics
Lecture 5: Agent-based modelling, Situation Awareness, Multi-Agent Situation Awareness, Differences in multi-agent situation awareness. Hands-on exercise: Evaluate a given accident scenario.
Lecture 6: Models of Human information processing, Human error, Error recovery, Cognitive control model, Complementary sub-models in agent-based modelling of hazards. Hands-on exercise: Continue evaluation of given accident scenario.
4.6月7日（星期四）: Petri net modelling syntax
Lecture 7: Explaining the formalisms of ordinary Petri nets, stochastic Petri nets and coloured Petri nets. Hands-on exercise: Apply the material to simple examples.
Lecture 8: Extensions of Petri nets to Stochastically and Dynamically Coloured Petri nets. Develop an agent-based model of an operation using Petri nets. Hands-on exercise: Apply the material to simple examples.
5.6月8日（星期五）: Simulation and validation
Lecture 9: From Petri net model specification to MC simulation. Accuracy of MC simulation results. Acceleration of MC simulation. Hands-on exercise: Reading a given Petri net specification.
Lecture 10: Model validation. Differences between model and reality. Using the MC simulation model for Uncertainty Quantification and Sensitivity Analysis. Hands-on exercise.
6.6月11日（星期一）: Q&A
Closing lecture: Benchmark of Agent-based versus Event-sequence based risk analysis. Questions and Answers.