Now open for registration....
International Symposium on Reliability Engineering and Risk Management (ISRERM)
|
De-Cheng Feng
|
Affiliation:
Title: Advanced Non-Parametric Fragility Analysis Method for Structures and Infrastructural Systems
Abstract: This keynote presentation introduces innovative non-parametric approaches to seismic fragility assessment of structures and infrastructural systems, which include three contents: (1) Copula-based full-probabilistic characterization: This method employs copula theory to parameterize the joint PDF of IM and EDP. Copulas provide a flexible and analytical framework for capturing dependencies between multiple variables, allowing us to derive joint PDF models tailored to specific engineering structures and seismic conditions. This approach bypasses the empirical assumptions of statistical moments and offers practical utility in engineering applications. The copula-based method enables precise modeling of joint PDF, facilitating high-fidelity seismic fragility curves. (2) Decoupled multi-probability density evolution method (M-PDEM): Building on the latest advancements in PDEM, the decoupled M-PDEM provides a non-parametric framework for quantifying the joint PDF of IM and EDP. This method involves solving a set of decoupled one-dimensional partial differential equations (PDEs) known as the Li-Chen equations. The equations govern the evolution of transient PDFs of the seismic-structural response quantities. By solving these decoupled equations, we can obtain the joint PDF of IM and EDP without resorting to high-dimensional PDEs, ensuring both computational efficiency and high numerical accuracy. The decoupled M-PDEM is physically driven, deriving the necessary joint PDF directly from representative analysis data, thus achieving superior accuracy over traditional empirical fitting techniques. (3) Poisson binomial distribution function for regional-scale assessment: This approach involves deriving the seismic fragility model of building portfolios at a regional scale using the Poisson binomial distribution. Individual buildings can be represented by simplified MDOF models that capture building-to-building variability, and these models can then be used to create a comprehensive dataset for training a probabilistic machine learning model. This model is capable of quantifying the inherent randomness in building responses and providing the probability distribution parameters needed for parametrized fragility models. By employing the Poisson binomial distribution, this method allows for assessing the seismic fragility of building portfolios rather than solely focusing on individual building damage probabilities as done in previous regional-scale research. These methods have been successfully applied to different structures and infrastructural systems, providing valuable insights into the resilience of complex engineering structures and representing a significant advancement in performance-based earthquake engineering (PBEE). By integrating sophisticated probabilistic modeling techniques and addressing the limitations of traditional empirical approaches, this method ensures high-accuracy fragility analysis, enhancing the seismic safety and resilience of complex engineering structures.
Keywords: Seismic fragility assessment; Full-probabilistic cloud analysis; Copulas; Decoupled multi-probability density evolution method (M-PDEM); Performance-based earthquake engineering (PBEE).
Personal Profile:
|
|
Matthias Faes
|
Affiliation:
Title:
Abstract: Uncertainties are especially commonly encountered in the context of structural dynamics, where for instance the effect natural phenomena such as earthquakes or wind loads on structures has to be considered. Indeed, due to the sheer complexity of the underlying physics, the corresponding dynamical loads that act on the system often cannot be described in a crisp way. Stochastic processes provide a rigorous framework to deal with the uncertainties and space/time correlations of uncertain loads by resorting to the well-documented framework of probability theory. However, in practice, the analyst is often confronted with limited, incomplete or conflicting sources of data (i.e., epistemic uncertainty). In this case, the application of a pure probabilistic framework to take this additional level of uncertainty into account is questionable since in this case, there is simply not enough information to construct an objective probabilistic uncertainty model. In this presentation, I will talk about how to deal with this challenging problem of modelling uncertainties in space and/or time under limited data. More precisely, I will show how to define and model imprecise stochastic processes that are robust with respect to missing and/or conflicting data, as well as present some efficient methodologies to effectively propagate these processes through numerical simulation models.
Personal Profile:
|
|
Michael Beer
|
Affiliation: Institute for Risk and Reliability, Leibniz University Hannover, Germany
Title:
Abstract: Engineered systems are most important for our daily life, they are the technical backbone of our society. A key requirement is, thus, to ensure their reliable performance. Reliability and performance analysis, however, become increasingly complicated due to uncertainties and complexity. This challenge needs to be addressed by a powerful modeling and analysis technology, which is able to deliver high-quality results quickly and to capture uncertainties comprehensively and realistically. The concept of survival signature provides a technology to summarize the systems availability in a condensed, statistical manner exploiting exchangeability of components of the same type. The systems reliability can then be formulated as a conditional probability, given that a certain number of components are still working, multiplied by the probability that this condition holds, which is controlled by the lifetime distributions of the components. Hence, the analysis of the system, to determine the conditional probability, is separated from the sampling on the lifetime distributions. The condensed systems representation, the survival signature, is used as a surrogate in the sampling approach, making it highly flexible and efficient. However, the determination of the survival signature remains as a combinatorial challenge with limitation to binary state components and systems. This challenge and limitation are addressed with efficient approximation schemes. Percolation theory is used to reduce complexity of the problem significantly by pre-elimination of irrelevant parts of the survival signature. Then, selected entries of the survival signature are estimated by Monte Carlo simulation, based on which a radial basis function network is trained to deliver an overall surrogate for the survival signature. To break the restriction of binary states, a continuous structure function is introduced. The resulting combinatorial problem is bypassed by approaching the solution from the side of the survival function instead from a system state analysis. A contour representation of the continuous-state survival function is combined with the concept of diagonally approximated signature to solve this problem. The lecture will guide the audience through this development. Illustrative engineering examples will be presented to demonstrate the capabilities of the approaches and concepts.
Personal Profile: Michael Beer is Professor and Head of the Institute for Risk and Reliability, Leibniz Universität Hannover, Germany. He is also part time Professor at the University of Liverpool and guest Professor at Tongji University and Tsinghua University, China. He obtained a doctoral degree from Technical University Dresden, Germany, and worked for Rice University, National University of Singapore, and the University of Liverpool, UK. Dr. Beer’s research is focused on uncertainty quantification in engineering with emphasis on imprecise probabilities. Dr. Beer is Editor in Chief of the ASCE-ASME Journal of Risk and Uncertainty in Engineering Systems, Part A Civil Engineering and Part B Mechanical Engineering. He is also Editor in Chief (joint) of the Encyclopedia of Earthquake Engineering, Associate Editor of Information Sciences, and Editorial Board Member of Engineering Structures and several other international journals. He has won several awards including the Alfredo Ang Award on Risk Analysis and Management of Civil Infrastructure of ASCE. Dr. Beer is the Chairman of the European Safety and Reliability Association (ESRA) and a Co-Chair of Risk and Resilience Measurements Committee (RRMC), Infrastructure Resilience Division (IRD), ASCE. He is serving on the Executive Board of the International Safety and Reliability Association (IASSAR), on the Executive Board of the European Association of Structural Dynamics (EASD), and on the Board of Directors of the International Association for Probabilistic Safety Assessment and Management (IAPSAM). He is a Fellow of the Alexander von Humboldt-Foundation and a Member of ASCE (EMI), ASME, CERRA, IACM and GACM.
|
|
Takeshi Kitahara
|
Affiliation: Department of Civil Engineering, Kanto Gakuin University, Yokohama, Japan
Title:
Abstract: Under circumstances where various uncertainties exist, such as the remarkable escalation of natural phenomena, including global warming, in recent years, there is an increasing need to consider events beyond the scope of assumption in the design of civil engineering structures. Here, an attempt to classify the relationship between assumed events in the design and various events beyond the scope of assumption based on the degree of perception of uncertainty is described. By rationally dividing these two phenomena in the design, it becomes possible to implement measures to increase robustness and redundancy on the boundary on a scenario basis as a structural plan. Furthermore, it is also expected to become feasible to clarify the positioning of anti-catastrophe, which encompasses the assumed events and the events beyond the scope of assumption.
Key words: event beyond scope of assumption, anti-catastrophe, uncertainty, cognition and knowledge
Personal Profile: Takeshi Kitahara is a Professor and Head of the Department of Civil Engineering at Kanto Gakuin University, Japan. He obtained a master’s degree from Kyoto University in 1991 and a doctorate from Nagoya University in 2001, Japan. He worked at the Research & Development Institute of Takenaka Corporation and Gunma National College of Technology. Dr. Kitahara’s research is focused on structural reliability, considering hybrid uncertainties in dynamic problems. He is also interested in applying AI and Data Science methods to a broader field of science. Dr. Kitahara is a Chairman of the JSCE Structural Safety Liaison Subcommittee. He is also a Chairman of the Japan Conference on Structural Safety and Reliability 2023 and was Editor in Chief of the Journal of Structural Engineering, JSCE. He is a member of JSCE, ASCE, ESRA, IABSE, IABMAS, IALCCE, and so on.
|
|
Suk Joo Bae
|
Affiliation: Department of Industrial Engineering, Hanyang University, Seoul, Korea
Title: Prognostic Health Management (PHM) and Predictive Maintenance (PdM) for Complex Systems
Abstract: Maintenance optimization for complex systems is an increasing critical issue in manufacturing industries including automobiles and semiconductors. Using IoT and smart censors, engineers aim to decide proper maintenance time points or intervals via health indicators representing system conditions. In this seminar, I introduce prognostic health management (PHM) and predictive maintenance (PdM) via off-line and on-line data for complex systems. Using off-line data, I present statistical models (e.g., nonhomogeneous Poisson process (NHPP), frailty models) for repairable systems. For PHM, I introduce a general five-stage process for PHM and PdM. I also present condition based maintenance policy using signal processing and statistical process control techniques, based on on-line sensor data. Finally, I present several real case studies for PHM and PdM in power plants and automobiles.
Personal Profile: Dr. Suk Joo Bae is a Professor in Hanyang University, Seoul, Republic of Korea. He was a Provost in Graduate School at Hanyang University, 2021-2023. Prof. Bae received his PhD. from the ISyE Department at Georgia Tech, 2003. He was the Editor-in-Chief of Journal of the Korean Society for Quality Management, The Journal of Applied Reliability, and the Associate Editor of IEEE Transactions on Reliability, and currently editorial board in ASCE-ASME Journal of Risk and Uncertainty in Engineering Systems, Part A: Civil Engineering from 2022. He was a President, Korean Society for Prognostics and Health Management (PHM), 2023. Prof. Bae has published more than 100 journal papers including Technometrics, Journal of Quality Technology, Reliability Engineering & System Safety, IISE Transactions, and IEEE Transactions on Reliability.
|
|
Yongbo Peng
|
Affiliation: Shanghai Instiute of Disaster Prevention and Relief, Tongji University, China
Title: Stochastic Optimal Control of Structures Based on Reliability Metrics
Abstract: Stochastic optimal control of structures presents a compelling approach to enhancing performance and fostering robust design in engineering structures and infrastructure systems. This methodology harnesses control and optimization techniques, accounting for stochastic dynamics inherent in natural hazards like earthquakes and typhoons, with a particular emphasis on ensuring structural safety grounded in reliability metrics. In this lecture, the historical trajectory of stochastic optimal control of structures is first sorted, and the basic formulas of the physically-based stochastic optimal (PSO) control are introduced. Subsequently, the generalized optimal control policy, encompassing the determination of control modalities, the selection of control devices, the definition of control laws, the design of critical control parameters, and the optimization of device layouts, for the PSO control is presented. Numerical examples of optimal structural control in active, semiactive, passive modalities showcase the advantages of the PSO control in bolstering structural safety and optimizing design expenses. The optimal stochastic compensation strategy for time-delayed feedback structural control systems is examined as well. Finally, a few practical applications are explored.
Personal Profile: Dr Yongbo Peng has been devoted in the area of uncertainty propagation and quantification, structural reliability and resilience analysis, stochastic control and optimization, and adaptive materials and structures. He obtained a PhD degree in Structural Engineering from Tongji University in 2009 after fulfilling a two-year joint PhD program in University of Southern California, USA, and joined Tongji University since then. He became a full Professor in 2018. Currently, he serves as a council member and deputy secretary-general of the Chinese Society for Vibration Engineering, a council member of the Shanghai Society of Theoretical and Applied Mechanics, and an editorial board member of the Journal of Architecture and Civil Engineering and Disaster Prevention and Resilience. Dr Peng has published two monographs, and over 130 journal papers. He won the Youth Science and Technology Award of the Chinese Society for Vibration Engineering in 2016, the second prize for technological invention and the first prize for natural science of the Chinese Society for Vibration Engineering in 2019 and 2023, and selected in the Stanford's list of top 2% most cited scientists in the world.
|
|
Yu Wang
|
Affiliation: Department of Architecture and Civil Engineering, City University of Hong Kong, Hong Kong, China
Title: Data-Driven Modelling and Simulation of Non-Gaussian Nonstationary Random Fields Directly from Sparse Measurements
Abstract: Modelling and simulation of random fields for spatially or temporally varying data play a pivotal role in reliability engineering and risk management, e.g., resilience assessment of civil infrastructures, particularly underground geo-structures. Data on civil infrastructures often vary spatially or temporally, and they might also be non-Gaussian and non-stationary. On the other hand, measurements on civil infrastructures, particularly subsurface space, are frequently sparse (e.g., at limited locations). It is therefore challenging to model or simulate the spatiotemporally varying, non-Gaussian, non-stationary data directly from sparse measurements. Data-driven methods are appealing to tackle this challenge because they bypass the difficulty in the selection of suitable parametric models or function forms and offer great flexibility for mimicking complicated characteristics of spatiotemporally varying data in a non-parametric manner. This talk presents some recent developments on random field modelling and simulation of spatiotemporally varying data under a data-driven framework of spectral representation or compressive sensing/sampling (CS). Similarity and differences between the spectral representation-based methods and the CS-based methods are discussed, including modelling of unknown trend function, marginal probability density function (Gaussian or non-Gaussian), and spatial or temporal autocovariance structure (stationary or non-stationary). Advantages of the CS-based methods are highlighted, such as superior performance for sparse measurements and incorporation of the uncertainty associated with the interpretation of sparse measurements. Personal Profile: Dr. Yu Wang is a professor of geotechnical engineering at City University of Hong Kong, and an elected Fellow of American Society of Civil Engineers (ASCE). His recent research efforts have focused on geotechnical uncertainty, reliability and risk, digital twin of subsurface geo-structures, machine learning in geotechnical engineering, analytics and simulation of geo-data, and seismic risk assessment of critical civil infrastructure systems. His research has earned several prestigious international/national awards, including the 2023 Thomas A. Middlebrooks Award from ASCE, the 2022 R.M. Quigley Award (Honourable Mention) from Canadian Geotechnical Society, the 2020 Best Paper Award from the ASCE-ASME Journal of Risk and Uncertainty in Engineering Systems, Part A: Civil Engineering, the 2020 Higher Education Outstanding Scientific Research Output Awards (the First-class Natural Science Award) from the Ministry of Education, China, the First-class Natural Science Award from the Hubei Provincial Government in 2017, the Highly Cited Research Award from the international journal of Engineering Geology in 2017, and the GEOSNet Young Researcher Award from the Geotechnical Safety Network (GEOSNet) in 2015. Dr Wang has authored/co-authored over 180 journal papers and two books in English. He served as president of ASCE Hong Kong Section in 2012-2013 and currently serves in editorial boards of several top journals in risk and uncertainty analysis or geotechnical engineering (e.g., Associate Editor for the ASCE Journal of Geotechnical and Geoenvironmental Engineering).
|
|
Colin Caprani
|
Affiliation: Department of Civil Engineering, Monash University, Australia
Title: Insightful Modelling: What Colour is Your Box?
Abstract: The last few years have seen a tremendous increase in the use of artificial intelligence and machine learning (AI/ML) approaches in structural engineering research. However, the intrinsic opacity of these "black box" models presents significant challenges in developing new insights, as they often obscure the underlying mechanisms. This lecture contrasts black box models with "white box" approaches, such as Bayesian hierarchical modeling, which prioritize transparency and interpretability by requiring explicit hypotheses. This contrast is illustrated through a case study on highway bridge traffic load modeling, where a Bayesian hierarchical model not only enhances predictive accuracy but also uncovers critical correlations and reduces uncertainties—insights that are difficult, if not impossible, to achieve with black box models. Overall, this talk advocates for the continued use of white box, or even grey box, models in structural engineering research, emphasizing their role in generating new knowledge and advancing the field beyond mere prediction.
Personal Profile: Dr. Colin Caprani, FIABSE, FIEI, FIEAust, FIStructE, is a leading structural engineering academic specializing in bridge safety and performance monitoring. With extensive experience in both academia and industry, he contributes significantly to the field through research, leadership roles, code development, and expert commentary on structural safety.
|
|
Pengfei Wei
|
Affiliation: School of Power and Energy, Northwestern Polytechnical University, China
Title:
Abstract: It has been a consensus in many areas that uncertainty quantification (UQ) is of fundamental importance in many engineering and theoretical areas. The primary focus of UQ is to develop a set of models/methods/algorithms for quantifying/reducing/neglecting/managing alternative sources and types of uncertainties which result in the disagreement between computational model predictions and real responses of a physical system, and further to make trustworthy analysis, decision and design under uncertainty. Despite the consensus on UQ methodology framework, reliable and efficient computational methods for several UQ subtasks are much less insufficiently developed, intrinsically resulted from the nested loops involved in these subtasks. This lecture presents a general Bayesian computational methodology framework for breaking these nested loops with a guarantee of numerical efficiency and global convergence, and extends it for addressing uncertainty propagation (forward UQ), probabilistic model updating and design optimization under uncertainty (inverse UQ). Applications of these developments to numerical and engineering examples are also presented.
Personal Profile: Dr. Wei, born in July 5, 1987, is an associate professor affiliated to the school of power and energy at Northwestern Polytechnical University, where he also commits as a doctoral advisor. He is a Humboldt fellow and has conducted collaborative research at Leibniz University Hannover for two years between 2018 and 2020. His research interests include reliability engineering, uncertainty quantification, key algorithms of digital twin, etc., with also specific focus on their applications to aero-engine. As principal investigators, he has ever hosted many projects including three granted by NSFC. He has also published more than 80 SCI-indexed academic journal papers, and has also been listed among world’s top 2% of scientists 2023 by Stanford. He serves as the associate editor of the ASCE-ASME Journal of Risk and Uncertainty in Engineering Systems, Part A: Civil Engineering, Part B: Mechanical Engineering, and the editor members of Reliability Engineering & System Safety (Elsevier) as well as Machine Learning for Computational Science and Engineering (Springer).
|
