一、课程目标 Mission
随着智能科技的迅速发展和工业系统的日益复杂，安全与风险问题已成为当今世界面临的重要挑战。为培养更多具有安全工程专业知识和实践经验的人才，安全与风险工程前沿课程应运而生。课程旨在提高学员对安全和风险管控的认识和理解，培养其解决问题能力和创新思维，从而提供更好的安全和风险管控解决方案。
本届安全与风险工程前沿暑期学校（2023年8月7日至11日）由博盈彩票主办。受邀专家均为安全工程、风险管理、可靠性工程、韧性工程领域的世界知名学者（包含美国工程院院士、加拿大工程院院士各1人、安全领域国际期刊主编4人次）。
课程将为安全科学与工程、应急管理科学与工程、力学、机械工程、航空航天、车辆工程及过程工业等专业的研究生提供领域前沿科学知识和相关研究方法，为从事科学研究打下基础。同时，本课程所提供的知识和方法也可为相关专业教师、研究人员和工程师提供参考。
With the rapid progress of intelligent science and technology and the increasing complexity of engineering systems, safety and risk has become one of the most important challenges the world facing today. To cultivate more talents with safety professional knowledge and practical experience, the Summer School on Advances of Safety and Risk Engineering has emerged as expected. It aims to improve students' understanding of safety and risk management, cultivate their problem-solving ability and innovative thinking, and provide sound safety and risk management solutions.
The Summer School on Advances of Safety and Risk Engineering is a graduate course event held by the School of Mechatronical Engineering, Beijing Institute of Technology (BIT).
The course will provide advanced scientific knowledge and relevant research methods for postgraduates majoring in safety science and engineering, emergency science and engineering, mechanics, mechanical engineering, aerospace, vehicle engineering, and process engineering, laying a foundation for their scientific research. Simultaneously, the knowledge and methods elaborated in this course can also be used for reference by teachers, researchers and engineers in related fields.
二、课程介绍 Syllabus
2023年安全与风险工程前沿暑期学校将开设以下课程，共32课时。课程开设时间为北京时间8月7日-8月11日。
The 2023 Summer School on Advances of Safety and Risk Engineering will offer the following courses with a total of 32 class hours. The course will be offered from August 7th to August 11th (Beijing Time).
三、相关信息 General Information
本课程招生对象为在校研究生、博士后、教职员工、研究人员、工业研发工程师。本届暑期学校将采用线上授课方式（具体上课方式另行通知），不收取任何费用。北理工在读研究生及准研究生按学校规定可获得相应的学分。
This course is open to current graduates, postdocs, faulty, professionals in research institutions, and industrial R&D engineers. Lectures will be presented online& offline. No fees will be charged. BIT graduate students who finish the program will be given credits.
四、报名方式 Registration
报名时间为即日起至2023年7月27日，请扫描二维码在线提交报名表，报名后可扫码加入微信群（任一微信群即可；不加微信群也可，相关通知会通过电子邮件同步告知）。如添加微信群有问题，请添加课程联系人微信，拉入群。将于8月1日之前发放录取通知。
Registration time is from today to July 27, 2023. Register online via the QR Code. After you register, please join WeChat group to keep in contact. Acceptance notices will be issued by August 1st.
|
|
|
课程报名表 |
课程微信群 |
|
五、联系方式 Contact Us
联系人：孟会行
联系电话/微信：18518125208
联系邮箱: huixing.meng@bit.edu.cn
For inquiries, please contact the program administrator.
Huixing Meng
Tel: (86) 18518125208
Email: huixing.meng@bit.edu.cn
六、课程简介 Course Content
1、安全与风险工程前沿
Advances of Safety and Risk Engineering
课程简介：主要介绍风险与安全工程基础知识、毒理学与职业卫生、定性与定量风险评估、泄漏建模等。同时，课程将论述风险的两个维度（后果和概率）、本质安全设计、故障建模、维护规划等。
The course mainly introduces the fundamentals of risk and safety engineering, toxicology and industrial hygiene, qualitative and quantitative risk assessment, source and release modelling. In the meanwhile, the course will discuss the two side of risk: consequence and probability, as well as inherently safer design, failure modelling, maintenance scheduling, etc.
2、系统安全工程的一种新范式需求
The Need for a new Paradigm in System Safety Engineering
课程简介：关于事故/损失原因的传统假设已经运用数百年。但是世界变化巨大，这些假设不再成立。事故复杂性的提升和产业新技术的出现，特别是计算机，形成了新的事故原因。传统的因果关系模型和相关工具在理解和预防这些新型损失方面效果不佳。该报告将解释为何需要新的范式，提出系统工程可为一个更适合当今世界的因果关系模型（STAMP模型）提供基础。
The traditional assumptions about the cause of accidents/losses have served us well for several hundred years. But the world has changed enough that these assumptions are no longer true. Previously unparalleled levels of complexity and new technology, particularly computers, have created new causes of accidents. Traditional models of causality and the tools based on them are not effective in understanding and preventing these new types of losses. In this talk, I will explain why something new is needed and suggest that systems theory can provide the basis for an expanded model of causality (called STAMP) that better fits today’s world.
3、智能系统人机界面的安全挑战
Safety and Security Challenges in AI-Human Interface
课程简介：在人工智能时代，智能人机界面正深刻影响着复杂系统安全。该报告将讨论智能系统人机界面中的安全和安保挑战。
In the era of artificial intelligence (AI), AI-Human Interface is deeply influencing the safety and security in complex systems. This talk will discuss the Safety and Security Challenges in AI-Human Interface.
4、区域风险与韧性分析概述
An overview of regional risk and resilience analysis
课程简介：维护、修理或更换现有的脆弱、缺陷和劣化结构和基础设施是一项重要投资。为有效分配有限资金，在决策过程使用先进的风险分析工具至关重要。报告论述区域风险和韧性分析的一般框架，介绍考虑多种灾害和不同基础设施及其劣化、相互依赖影响的区域风险和韧性分析方法。报告还展示如何将结构和基础设施的物理损坏综合起来预测业务中断的可能性和持续时间，最后给出新马德里地震区一次假想地震的区域风险和韧性分析实例。
The maintenance, repair, or replacement of existing vulnerable, deficient, and deteriorating structures and infrastructure represents a significant investment. To wisely invest the limited funding, it is crucial to use advanced risk analysis tools in the decision-making process. This presentation discusses a general formulation for regional risk and resilience analysis. The presentation explains how to conduct a regional risk and resilience analysis considering multiple hazards and different infrastructure, as well as the effects of deterioration and interdependencies among infrastructure. The presentation also shows how the physical damage to structures and infrastructure can be cascaded to predict the likelihood and duration of business interruption. The presentation concludes with an example of regional risk and resilience analysis considering a hypothetical earthquake in the New Madrid seismic zone.
5、货运系统的数据驱动风险分析
Date-driven risk analysis in freight transport systems
课程简介：应对不同特征的风险需要整合跨学科、跨研究方法。目前我们缺乏对哪些类型的风险分析方案能够最有效地利用科学与技术实现长期弹性的、可持续的货物运输系统的关键理解。本报告旨在介绍系列数据驱动风险分析研究，将来自不同领域的风险信息量化、整合和交互，并促进风险文化从传统的反应性单维模式向新的主动性多维模式转变。这项工作将解决与韧性和可持续性科学相关的重要方法学问题，特别是在高度不确定性下的风险分析。
Addressing risks of different features requires integration across disciplines and across research methodologies. Currently, we lack the critical understanding of which kinds of risk schemes can most effectively harness science and technology for achieving long-term resilient and sustainable freight transport systems (FTSs). This presentation aims at introducing a series of studies on data driven risk analysis which enables the quantification, integration and communication of risk information from different areas and facilitates the movement of risk culture from a traditional reactive single dimensional scheme towards a new proactive multiple dimensional regime. The work will address the significant methodological issues associated with resilience and sustainability sciences particularly with reference to risk analysis under high uncertainty.
6、相依失效系统的可靠性与维护分析
Reliability and maintenance analysis of systems with dependent failures
课程简介：随着技术的进步，由互联设备组成的现代系统变得日益复杂。不同组件之间的相互作用和依赖关系可能引发故障。从概率论角度来看，当其他构件状态变化时，该构件的失效概率可能发生变化，这种失效被称为相依失效。在本报告中将概述相依失效的类型以及近年关于相依失效系统的可靠性和维护分析的研究。
As technology advances, modern systems composed of interconnected devices become increasingly complicated. Failures can occur due to the interactions and dependences between different components. From the perspective of probability theory, the failure probability of one component can be different when the conditions of other components change, and such a failure is called as a dependent failure. In the talk, we will have an overview on the types of dependent failures and some research in these year on reliability and maintenance analysis of systems with dependent failures that the speaker participated in these years.
7、面向基础设施网络韧性增强的稳健可扩展的决策分析
Robust and scalable prescriptive analytics for infrastructure network resilience enhancement
课程简介：报告讨论决策分析在基础设施网络韧性增强方面的应用，重点讨论两个方面。首先，引入分布鲁棒优化作为模糊失效场景下的事前规划策略，展示典型案例研究证明方法有效性；其次，深入研究使用深度强化学习改善大规模现实世界系统的事后响应和恢复规划，讨论深度不确定性和计算复杂性带来的挑战，并对未来的研究方向进行展望。
This talk discusses the application of prescriptive analytics in enhancing infrastructure network resilience, with a focus on two key aspects. First, the seminar will introduce distributionally robust optimization as a strategy for ex-ante planning under ambiguous failure scenarios. The speaker will present exemplary case studies to demonstrate the efficacy of this approach. Second, the seminar will delve into the use of deep reinforcement learning to improve ex-post response and recovery planning for large-scale real-world systems. The speaker will discuss the challenges posed by deep uncertainties and computational complexity, and provide insights into future research directions.
