Data-driven uncertainty quantification and propagation in structural dynamics through a hierarchical Bayesian framework

In the presence of modeling errors, the mainstream Bayesian methods seldom give a realistic account of uncertainties as they commonly underestimate the inherent variability of parameters. This problem is not due to any misconceptions in the Bayesian framework since it is robust with respect to the modeling assumptions and the observed data. Rather, this issue has deep roots in users’ inability to develop an appropriate class of probabilistic models. This paper bridges this significant gap, introducing a novel Bayesian hierarchical setting, which breaks time-history vibration responses into several segments so as to capture and identify the variability of inferred parameters over the segments. Since the computation of the posterior distributions in hierarchical models is expensive and cumbersome, novel marginalization strategies, asymptotic approximations, and maximum a posteriori estimations are proposed and outlined in a computational algorithm aiming to handle both uncertainty quantification and propagation. For the first time, the connection between the ensemble covariance matrix and hyper distribution parameters is characterized through approximate estimations. Experimental and numerical examples are employed to illustrate the efficacy and efficiency of the proposed method. It is observed that, when the segments correspond to various system operating conditions and input characteristics, the proposed method delivers robust parametric uncertainties with respect to unknown phenomena such as ambient conditions, input characteristics, and environmental factors.     

Inference and uncertainty quantification of stochastic gene expression via synthetic models

Estimating uncertainty in model predictions is a central task in quantitative biology. Biological models at the single-cell level are intrinsically stochastic and nonlinear, creating formidable challenges for their statistical estimation which inevitably has to rely on approximations that trade accuracy for tractability. Despite intensive interest, a sweet spot in this trade-off has not been found yet. We propose a flexible procedure for uncertainty quantification in a wide class of reaction networks describing stochastic gene expression including those with feedback. The method is based on creating a tractable coarse-graining of the model that is learned from simulations, a synthetic model, to approximate the likelihood function. We demonstrate that synthetic models can substantially outperform state-of-the-art approaches on a number of non-trivial systems and datasets, yielding an accurate and computationally viable solution to uncertainty quantification in stochastic models of gene expression.     

Quantifying uncertainty under a predictive, epistemic approach to risk analysis

Risk analysis is a tool for investigating and reducing uncertainty related to outcomes of future activities. Probabilities are key elements in risk analysis, but confusion about interpretation and use of probabilities often weakens the message from the analyses. Under the predictive, epistemic approach to risk analysis, probabilities are used to express uncertainty related to future values of observable quantities like the number of fatalities or monetary loss in a period of time. The procedure for quantifying this uncertainty in terms of probabilities is, however, not obvious. Examples of topics from the literature relevant in this discussion are use of expert judgement, the effect of so-called heuristics and biases, application of historical data, dependency and updating of probabilities. The purpose of this paper is to discuss and give guidelines on how to quantify uncertainty in the perspective of these topics. Emphasis is on the use of models and assessment of uncertainties of similar quantities.     

Efficient uncertainty quantification of dynamical systems with local nonlinearities and uncertainties

Idealized modeling of most engineering structures yields linear mathematical models, i.e., linear ordinary or partial differential equations. However, features like nonlinear dampers and/or springs can render nonlinear an otherwise linear model. Often, the connectivity of these nonlinear elements is confined to only a few degrees-of-freedom (DOFs) of the structure. In such cases, treating the entire structure as nonlinear results in very computationally expensive solutions. Moreover, if system parameters are uncertain, their stochastic nature can render the analysis even more computationally costly. This paper presents an approach for computing the response of such systems in a very efficient manner. The proposed solution procedure first segregates the DOFs appearing in the nonlinear and/or stochastic terms from those DOFs that involve only linear deterministic operations. Second, the responses of nonlinear/stochastic terms are determined using a non-standard form of a nonlinear Volterra integral equation (NVIE). Finally, the responses of the remaining DOFs are computed through a convolution approach using the fast Fourier transform to further increase the computational efficiency. Three examples are presented to demonstrate the efficacy and accuracy of the proposed method. It is shown that, even for moderately sized systems (∼1000 DOFs), the proposed method is about three orders of magnitude faster than a conventional Monte Carlo sampling method (i.e., solving the system of ODEs repeatedly).     

Gated Linear Model induced U-net for surrogate modeling and uncertainty quantification

We propose a novel deep learning based surrogate model for solving high-dimensional uncertainty quantification and uncertainty propagation problems. The proposed deep learning architecture is developed by integrating the well-known U-net architecture with the Gaussian Gated Linear Network (GGLN) and referred to as the Gated Linear Network induced U-net or GLU-net. The proposed GLU-net treats the uncertainty propagation problem as an image to image regression and hence, is extremely data efficient. Additionally, it also provides estimates of the predictive uncertainty. The network architecture of GLU-net is less complex with 44% fewer parameters than the contemporary works. We illustrate the performance of the proposed GLU-net in solving the Darcy flow problem under uncertainty under the sparse data scenario. We consider the stochastic input dimensionality to be up to 4225. Benchmark results are generated using the vanilla Monte Carlo simulation. We observe the proposed GLU-net to be accurate and extremely efficient even when no information about the structure of the inputs is provided to the network. Case studies are performed by varying the training sample size and stochastic input dimensionality to illustrate the robustness of the proposed approach.     

Bayesian analysis of computer code outputs: A tutorial

The Bayesian approach to quantifying, analysing and reducing uncertainty in the application of complex process models is attracting increasing attention amongst users of such models. The range and power of the Bayesian methods is growing and there is already a sizeable literature on these methods. However, most of it is in specialist statistical journals. The purpose of this tutorial is to introduce the more general reader to the Bayesian approach.     

Physics-integrated deep learning for uncertainty quantification and reliability estimation of nonlinear dynamical systems

Assumptions and approximations made while analyzing any physical system induce modeling uncertainty, which, if left unchecked, can result in the erroneous analysis of the system under consideration. Additionally, the discrepancy in the exact knowledge of system parameters can further result in deviation from the ground truth. This paper explores Physics-integrated Variational Auto-Encoder (PVAE) to account for modeling and parametric uncertainties in partially known nonlinear dynamical systems. The PVAE under consideration has three main parts: encoder, latent space, and decoder. The complete PVAE architecture is employed during the training stage of the machine learning model, while only the decoder is used to make the final predictions. The encoder determines the correct parameter values for the known part of the model (in the form of a known ODE). The decoder is augmented with an ODE solver that solves the known part of the system and the estimated discrepancy together to reconstruct the measurements. To test the efficacy of the PVAE architecture, three case studies are carried out, each presenting unique challenges. The probability density functions obtained for the various systems’ responses demonstrate the efficacy of the PVAE architecture. Furthermore, reliability analysis has been carried out, and the results produced have been compared against those obtained from a multi-layered, densely connected forward neural network.     

Comparison of Bayesian methods on parameter identification for a viscoplastic model with damage

The state of materials and accordingly the properties of structures are changing over the period of use, which may influence the reliability and quality of the structure during its life-time. Therefore, identification of the model parameters of the system is a topic which has attracted attention in the content of structural health monitoring. The parameters of a constitutive model are usually identified by minimization of the difference between model response and experimental data. However, the measurement errors and differences in the specimens lead to deviations in the determined parameters. In this article, the focus is on the identification of material parameters of a viscoplastic damaging material using a stochastic simulation technique to generate artificial data which exhibit the same stochastic behavior as experimental data. It is proposed to use Bayesian inverse methods for parameter identification and therefore the model and damage parameters are identified by applying the Transitional Markov Chain Monte Carlo Method (TMCMC) and Gauss–Markov–Kalman filter (GMKF) approach. Identified parameters by using these two Bayesian approaches are compared with the true parameters in the simulation and with each other, and the efficiency of the identification methods is discussed. The aim of this study is to observe which one of the mentioned methods is more suitable and efficient to identify the model and damage parameters of a material model, as a highly non-linear model, using a limited surface displacement measurement vector and see how much information is indeed needed to estimate the parameters accurately.     

动态有效位数的测量不确定度

介绍了四参数正弦波最小二乘法评价动态有效位数指标，以及其不确定度分析和评价过程，同时给出了一个分析评价实例，该过程及结论可应用在对于测量标准进行相应指标的不确定度分析上，也可用于估计指标本身的不确定性。     

Non‐linear model reduction for uncertainty quantification in large‐scale inverse problems

We present a model reduction approach to the solution of large‐scale statistical inverse problems in a Bayesian inference setting. A key to the model reduction is an efficient representation of the non‐linear terms in the reduced model. To achieve this, we present a formulation that employs masked projection of the discrete equations; that is, we compute an approximation of the non‐linear term using a select subset of interpolation points. Further, through this formulation we show similarities among the existing techniques of gappy proper orthogonal decomposition, missing point estimation, and empirical interpolation via coefficient‐function approximation. The resulting model reduction methodology is applied to a highly non‐linear combustion problem governed by an advection–diffusion‐reaction partial differential equation (PDE). Our reduced model is used as a surrogate for a finite element discretization of the non‐linear PDE within the Markov chain Monte Carlo sampling employed by the Bayesian inference approach. In two spatial dimensions, we show that this approach yields accurate results while reducing the computational cost by several orders of magnitude. For the full three‐dimensional problem, a forward solve using a reduced model that has high fidelity over the input parameter space is more than two million times faster than the full‐order finite element model, making tractable the solution of the statistical inverse problem that would otherwise require many years of CPU time. Copyright © 2009 John Wiley & Sons, Ltd.     

新型直流电阻箱检定装置的不确定度评定

采用上海蓝斯汀仪表研究所生产的RT9608直流电阻校验仪和一套二等标准电阻,对ZX54型0.01级直流电阻箱×1000档第一点的测量结果的不确定度评定过程的阐述与分析,指出了影响测量结果的相关因素,并通过计算合成进行了评定。结果证明,该新型装置完全符合规程要求,自动化程度高,具有很广的应用前景。     

音速喷嘴气体流量标准装置测量不确定度评定

各种流量计从研制到使用，必须对其特性进行研究，对其标称的流量性能进行试验，并给出试验结果。这些试验研究工作都需要在流量标准装置上进行。因此，流量标准装置的不确定度直接关系到试验结果的水平，对其研究是非常必要的。用临界流文丘利喷嘴做为流量标准，用比较法原理对流量计进行检定和校准，音速喷嘴法气体流量标准装置可对各种气体流量计的检测和校准，并对其检定和校准结果的不确定度进行评定。     

校准喷嘴系统的不确定度分析方法

介绍了热线／热膜流速仪的基本原理和用于热线／热膜流速仪校准用的校准喷嘴系统的气体速度确定方法，分析和推导了用不确定度理论全面分析系统不确定度的方法。     

非接触式汽车速度计校准装置研制

本文介绍了非接触式汽车速度计的作用及检测原理，详细叙述它的主要技术要求、现有的校准方法以及研制作为国内首创的新型校准装置的工作原理、技术参数、测量不确定度及实际校准非接触式汽车速度计的实验结果。目前已开展对非接触式汽车速度计的校准工作。     

Framework for sensitivity and uncertainty quantification in the flutter assessment of bridges

The phenomenon of aerodynamic instability caused by wind is usually a major design criterion for long-span cable-supported bridges. If the wind speed exceeds the critical flutter speed of the bridge, this constitutes an Ultimate Limit State. The prediction of the flutter boundary therefore requires accurate and robust models. The state-of-the-art theory concerning determination of the flutter stability limit is presented. Usually bridge decks are bluff and therefore the aeroelastic forces under wind action have to be experimentally evaluated in wind tunnels or numerically computed through Computational Fluid Dynamics (CFD) simulations. The self-excited forces are modelled using aerodynamic derivatives obtained through CFD forced vibration simulations on a section model. The two-degree-of-freedom flutter limit is computed by solving the Eigenvalue problem.A probabilistic flutter analysis utilizing a meta-modelling technique is used to evaluate the effect of parameter uncertainty. A bridge section is numerically modelled in the CFD simulations. Here flutter derivatives are considered as random variables. A methodology for carrying out sensitivity analysis of the flutter phenomenon is developed. The sensitivity with respect to the uncertainty of flutter derivatives and structural parameters is considered by taking into account the probability distribution of the flutter limit. A significant influence on the flutter limit is found by including uncertainties of the flutter derivatives due to different interpretations of scatter in the CFD simulations. The results indicate that the proposed probabilistic flutter analysis provides extended information concerning the accuracy in the prediction of flutter limits.The final aim is to set up a method to estimate the flutter limit with probabilistic input parameters. Such a tool could be useful for bridge engineers at early design stages. This study shows the difficulties in this regard which have to be overcome but also highlights some interesting and promising results.     

The use of global uncertainty methods for the evaluation of combustion mechanisms

Complex chemical mechanisms are increasingly used within models describing a range of important chemical processes. Parameters describing the rates of chemical steps and thermodynamics may be highly uncertain, influencing the uncertainty in final model predictions. Local sensitivity analysis is traditionally employed within commercial modelling packages but may not be appropriate for highly uncertain data within non-linear models. There is a need for global uncertainty techniques such as Morris and Monte Carlo methods that can be applied efficiently for computationally expensive models. This paper presents the development of such techniques, along with application to a kinetic mechanism describing the influence of fuel trace elements such as sulphur-containing compounds, on the formation of nitrogen oxide in combustion devices. The analysis evaluates the parameters from within the current sulphur scheme that drive uncertainties in predicted relative changes in nitrogen oxide concentrations when sulphur compounds are added to the fuel. The overall performance of the mechanism is evaluated in comparison with available experimental profiles and the level of agreement between different methods for importance ranking of the rate parameters is highlighted. The use of fitted model representations is also discussed as an alternative method for determining importance ranking, and highlights non-linear interactions between parameters. Finally, possible improvements to the chemical scheme are tested within a Monte Carlo framework under lean flame conditions, where the current mechanism performs the least well with respect to experimental results.     

An efficient metamodeling approach for uncertainty quantification of complex systems with arbitrary parameter probability distributions

This paper proposes an efficient metamodeling approach for uncertainty quantification of complex system based on Gaussian process model (GPM). The proposed GPM‐based method is able to efficiently and accurately calculate the mean and variance of model outputs with uncertain parameters specified by arbitrary probability distributions. Because of the use of GPM, the closed form expressions of mean and variance can be derived by decomposing high‐dimensional integrals into one‐dimensional integrals. This paper details on how to efficiently compute the one‐dimensional integrals. When the parameters are either uniformly or normally distributed, the one‐dimensional integrals can be analytically evaluated, while when parameters do not follow normal or uniform distributions, this paper adopts the effective Gaussian quadrature technique for the fast computation of the one‐dimensional integrals. As a result, the developed GPM method is able to calculate mean and variance of model outputs in an efficient manner independent of parameter distributions. The proposed GPM method is applied to a collection of examples. And its accuracy and efficiency is compared with Monte Carlo simulation, which is used as benchmark solution. Results show that the proposed GPM method is feasible and reliable for efficient uncertainty quantification of complex systems in terms of the computational accuracy and efficiency. Copyright © 2016 John Wiley & Sons, Ltd.     

A Bayesian framework for uncertainty quantification of perturbed gamma process based on simulated likelihood

The perturbed gamma process (PGP) has recently been widely used in modeling the noisy degradation data collected from engineering structures and components since it can simultaneously consider the temporal variability of degradation and measurement uncertainty. As a result of the sampling and inspection uncertainty in engineering practice, it is necessary to account for the resulting parameter uncertainty. Meanwhile, the flexibility of the form of measurement error motivates a potential demand for quantifying the model uncertainty and selecting the most fitting error model for the given inspection data. The Bayesian approach is well-suited to quantity the parameter uncertainty induced by imperfect inspection and limited inspection data, but its practical implementation is extremely challenging due to the intractable likelihood function of PGP. In the paper, a novel Bayesian framework for quantifying parameter and model uncertainty of PGP is presented, where the simulated likelihood that is an unbiased estimator generated by Sequential Monte Carlo (SMC) is introduced to overcome the intractable likelihood of PGP. More specifically, an Adaptive Particle Markov chain Monte Carlo (APMCMC) is proposed to perform the Bayesian sampling from the posterior distributions of parameters, achieving the requirement for the quantification of parameter uncertainty. By utilizing the posterior samples from APMCMC, a model selection method based on the Bayes factor is employed to determine the most fitting one from some alternative error models. Finally, two simulation examples are presented to illustrate the efficiency and accuracy of the proposed framework and its applicability is confirmed by a practical case involving the corrosion modeling of a group of pipelines.     

血细胞分析仪检定(校准)测量结果的不确定度评定

通过血细胞分析仪检定／校准装置的测量结果不确定度评定，确定检定／校准装置满足被检血细胞分析仪检定／校准的要求。可供需要建立血细胞分析仪检定／校准装置的相关单位中有关人员进行不确定度分析和计算时参考。