Browsing by Author "Li, Huiru"

Now showing 1 - 4 of 4
  • Thumbnail Image
    Item
    A Bayesian Approach to Recovering Missing Component Dependence for System Reliability Prediction via Synergy Between Physics and Data
    (ASME, 2021-11) Li, Huiru; Du, Xiaoping; Mechanical and Energy Engineering, School of Engineering and Technology
    Predicting system reliability is often a core task in systems design. System reliability depends on component reliability and dependence of components. Component reliability can be predicted with a physics-based approach if the associated physical models are available. If the models do not exist, component reliability may be estimated from data. When both types of components coexist, their dependence is often unknown, and the component states are therefore assumed independent by the traditional method, which can result in a large error. This work proposes a new system reliability method to recover the missing component dependence, thereby leading to a more accurate estimate of the joint probability density (PDF) of all the component states. The method works for series systems whose load is shared by its components that may fail due to excessive loading. For components without physical models available, the load data are recorded upon failure, and equivalent physical models are created; the model parameters are estimated by the proposed Bayesian approach. Then models of all component states become available, and the dependence of component states, as well as their joint PDF, can be estimated. Four examples are used to evaluate the proposed method, and the results indicate that the proposed method can produce more accurate predictions of system reliability than the traditional method that assumes independent component states.
  • Thumbnail Image
    Item
    Label Free Uncertainty Quantification
    (ARC, 2022-01) Li, Huiru; Yin, Jianhua; Du, Xiaoping; Mechanical Engineering, School of Engineering and Technology
    View Video Presentation: https://doi.org/10.2514/6.2022-1097.vid Uncertainty quantification (UQ) is essential in scientific computation since it can provide the estimate of the uncertainty in the model prediction. Intensive computation is required for UQ as it calls the deterministic simulation repeatedly. This study discusses a physics-based label-free deep learning UQ method that does not need predictions at training points or labels. It satisfies the physical equations from which labels could be generated without solving the equations during the training process. Then inexpensive surrogate models are built with respect to model inputs. The surrogate models are used for UQ with a much lower computational cost. Two examples demonstrate that the label-free method can efficiently produce probability distributions of model outputs for given distributions of random input variables.
  • Thumbnail Image
    Item
    Physics-Based Regression vs. CFD for Hagen-Poiseuille and Womersley Flows and Uncertainty Quantification
    (NSF-PAR, 2022-07) Islam, Md Mahfuzul; Li, Huiru; Yu, Huidan; Du, Xiaoping; Surgery, School of Medicine
    Computational fluid dynamics (CFD) and its uncertainty quantification are computationally expensive. We use Gaussian Process (GP) methods to demonstrate that machine learning can build efficient and accurate surrogate models to replace CFD simulations with significantly reduced computational cost without compromising the physical accuracy. We also demonstrate that both epistemic uncertainty (machine learning model uncertainty) and aleatory uncertainty (randomness in the inputs of CFD) can be accommodated when the machine learning model is used to reveal fluid dynamics. The demonstration is performed by applying simulation of Hagen-Poiseuille and Womersley flows that involve spatial and spatial-tempo responses, respectively. Training points are generated by using the analytical solutions with evenly discretized spatial or spatial-temporal variables. Then GP surrogate models are built using supervised machine learning regression. The error of the GP model is quantified by the estimated epistemic uncertainty. The results are compared with those from GPU-accelerated volumetric lattice Boltzmann simulations. The results indicate that surrogate models can produce accurate fluid dynamics (without CFD simulations) with quantified uncertainty when both epistemic and aleatory uncertainties exist.
  • Thumbnail Image
    Item
    Recovering Missing Component Dependence for System Reliability Prediction via Synergy between Physics and Data
    (American Society of Mechanical Engineers, 2021) Li, Huiru; Du, Xiaoping; Mechanical and Energy Engineering, Purdue School of Engineering and Technology
    Predicting system reliability is often a core task in systems design. System reliability depends on component reliability and dependence of components. Component reliability can be predicted with a physics-based approach if the associated physical models are available. If the models do not exist, component reliability may be estimated from data. When both types of components coexist, their dependence is often unknown, and therefore, the component states are assumed independent by the traditional method, which can result in a large error. This study proposes a new system reliability method to recover the missing component dependence, thereby leading to a more accurate estimate of the joint probability density function (PDF) of all the component states. The method works for series systems whose load is shared by its components that may fail due to excessive loading. For components without physical models available, the load data are recorded upon failure, and equivalent physical models are created; the model parameters are estimated by the proposed Bayesian approach. Then models of all component states become available, and the dependence of component states, as well as their joint PDF, can be estimated. Four examples are used to evaluate the proposed method, and the results indicate that the method can produce more accurate predictions of system reliability than the traditional method that assumes independent component states.