Engineering Papers⌕ Search

Engineering topics

Hu, Zhen

Publications and source records attributed to Hu, Zhen.

At least 19 records
All sourcesNASA NTRSDOE OSTI
DOE OSTI2024

Probabilistic machine learning for battery health diagnostics and prognostics—review and perspectives

Abstract Diagnosing lithium-ion battery health and predicting future degradation is essential for driving design improvements in the laboratory and ensuring safe and reliable operation over a product’s expected lifetime. However, accurate battery health diagnostics and prognostics is challenging due to the unavoidable influence of cell-to-cell manufacturing variability and time-varying operating circumstances experienced in the field. Machine learning approaches informed by simulation, experiment, and field data show enormous promise to predict the evolution of battery health with use; however, until recently, the research community has focused on deterministic modeling methods, largely ignoring the cell-to-cell performance and aging variability inherent to all batteries. To truly make informed decisions regarding battery design in the lab or control strategies for the field, it is critical to characterize the uncertainty in a model’s predictions. After providing an overview of lithium-ion battery degradation, this paper reviews the current state-of-the-art probabilistic machine learning models for health diagnostics and prognostics. Details of the various methods, their advantages, and limitations are discussed in detail with a primary focus on probabilistic machine learning and uncertainty quantification. Last, future trends and opportunities for research and development are discussed.

25 ENERGY STORAGE↗
DOE OSTI2023

A global significance evaluation method using simulated events

In High-Energy Physics experiments it is often necessary to evaluate the global statistical significance of apparent resonances observed in invariant mass spectra. One approach to determining significance is to use simulated events to find the probability of a random fluctuation in the background mimicking a real signal. As a high school summer project, we demonstrate a method with Monte Carlo simulated events to evaluate the global significance of a potential resonance with some assumptions. This method for determining significance is general and can be applied, with appropriate modification, to other resonances.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2023

Uncertainty quantification in machine learning for engineering design and health prognostics: A tutorial

On top of machine learning (ML) models, uncertainty quantification (UQ) functions as an essential layer of safety assurance that could lead to more principled decision making by enabling sound risk assessment and management. The safety and reliability improvement of ML models empowered by UQ has the potential to significantly facilitate the broad adoption of ML solutions in high-stakes decision settings, such as healthcare, manufacturing, and aviation, to name a few. In this tutorial, we aim to provide a holistic lens on emerging UQ methods for ML models with a particular focus on neural networks and the applications of these UQ methods in tackling engineering design as well as prognostics and health management problems. Towards this goal, we start with a comprehensive classification of uncertainty types, sources, and causes pertaining to UQ of ML models. Next, we provide a tutorial-style description of several state-of-the-art UQ methods: Gaussian process regression, Bayesian neural network, neural network ensemble, and deterministic UQ methods focusing on spectral-normalized neural Gaussian process. Established upon the mathematical formulations, we subsequently examine the soundness of these UQ methods quantitatively and qualitatively (by a toy regression example) to examine their strengths and shortcomings from different dimensions. Then, we review quantitative metrics commonly used to assess the quality of predictive uncertainty in classification and regression problems. Afterward, we discuss the increasingly important role of UQ of ML models in solving challenging problems in engineering design and health prognostics. In conclusion, two case studies with source codes available on GitHub are used to demonstrate these UQ methods and compare their performance in the life prediction of lithium-ion batteries at the early stage (case study 1) and the remaining useful life prediction of turbofan engines (case study 2).

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2022

Charged particle tracking in real-time using a full-mesh data delivery architecture and associative memory techniques

We present a flexible and scalable approach to address the challenges of charged particle track reconstruction in real-time event filters (Level-1 triggers) in collider physics experiments. The method described here is based on a full-mesh architecture for data distribution and relies on the Associative Memory approach to implement a pattern recognition algorithm that quickly identifies and organizes hits associated to trajectories of particles originating from particle collisions. We describe a successful implementation of a demonstration system composed of several innovative hardware and algorithmic elements. The implementation of a full-size system relies on the assumption that an Associative Memory device with the sufficient pattern density becomes available in the future, either through a dedicated ASIC or a modern FPGA. We demonstrate excellent performance in terms of track reconstruction efficiency, purity, momentum resolution, and processing time measured with data from a simulated LHC-like tracking detector.

47 OTHER INSTRUMENTATION↗
DOE OSTI2022

Bayesian model updating with finite element vs surrogate models: Application to a miter gate structural system

Bayesian finite element (FE) model updating using direct model evaluations of large-scale high-fidelity FE models is extremely computationally expensive. Surrogate models can be used as fast emulators of FE models to accelerate the model calibration process. The physics/mechanics-based FE models are still the underpinning behind the surrogate models. Here, this paper evaluates the loss in accuracy and the gain in computational time while performing Bayesian model updating by using surrogate model evaluations compared to using direct FE model evaluations. This evaluation is crucial before entirely relying on surrogate models in model updating for structural health monitoring (SHM) and damage prognosis (DP) purposes. This paper also demonstrates Bayesian updating and surrogate model construction of large-scale high-fidelity FE models of infrastructure systems. In this regard, the miter gate structural system is considered as the testbed structure. Three predominant damage modes (loss of contact between gate and wall, loss of thickness due to corrosion, and loss of tension in the diagonal rods) are considered for model updating purposes. Bayesian model updating is performed using direct FE evaluations by leveraging parallel computing. Two types of surrogates, namely polynomial chaos expansion (PCE) and Gaussian process regression (GPR), are developed for the miter gate. Model updating is performed again using the trained surrogate models, and the updating results are compared with their counterparts obtained using the direct FE evaluation results. The posterior distribution of the FE model parameters obtained using the trained surrogates are sufficiently accurate with respect to the posterior obtained utilizing the direct FE evaluations. In addition, an approximate 4-fold decrease in the computational time was observed when using surrogate model evaluations instead of direct FE evaluations for model updating.

42 ENGINEERING↗
DOE OSTI2022

Surrogate Modeling of Nonlinear Dynamic Systems: A Comparative Study

Surrogate models play a vital role in overcoming the computational challenge in designing and analyzing nonlinear dynamic systems, especially in the presence of uncertainty. This paper presents a comparative study of different surrogate modeling techniques for nonlinear dynamic systems. Four surrogate modeling methods, namely, Gaussian process (GP) regression, a long short-term memory (LSTM) network, a convolutional neural network (CNN) with LSTM (CNN-LSTM), and a CNN with bidirectional LSTM (CNN-BLSTM), are studied and compared. All these model types can predict the future behavior of dynamic systems over long periods based on training data from relatively short periods. The multi-dimensional inputs of surrogate models are organized in a nonlinear autoregressive exogenous model (NARX) scheme to enable recursive prediction over long periods, where current predictions replace inputs from the previous time window. Three numerical examples, including one mathematical example and two nonlinear engineering analysis models, are used to compare the performance of the four surrogate modeling techniques. The results show that the GP-NARX surrogate model tends to have more stable performance than the other three deep learning (DL)-based methods for the three particular examples studied. The tuning effort of GP-NARX is also much lower than its deep learning-based counterparts.

42 ENGINEERING↗
DOE OSTI2021

Model correction and updating of a stochastic degradation model for failure prognostics of miter gates

Understanding the degradation of the quoin block is vital for failure prognostics in miter gates. Due to the complicated degradation mechanism, degradation models based on simplifications and assumptions cannot accurately describe the damage evolution. It is observed that small errors in a simplified degradation model can lead to a large discrepancy in the remaining useful life estimation attributed to error accumulation over time. Aiming to address this issue in failure prognostics, this paper presents a dynamic model correction framework for a simplified degradation model using strain measurements. In the proposed framework, a polynomial chaos expansion (PCE) model is employed to compensate the missing physics in a simplified stochastic degradation model. Here, a maximum likelihood estimation method is developed to estimate the uncertain parameters of the simplified physics-based degradation model along with the unknown PCE model parameters using strain measurements as the observables. The updated damage degradation model is then applied to failure prognostics of a miter gate. Results of a case study show that the proposed approach can effectively improve the accuracy of failure prognostics in miter gates.

42 ENGINEERING↗
DOE OSTI2020

Forward Physics Facility - Snowmass 2021 Letter of Interest

A rich physics program remains unexplored in the far-forward region at the LHC. The Forward Physics Facility (FPF) is a proposal to enlarge an existing cavern in the far-forward region of ATLAS to house a suite of experiments with groundbreaking new capabilities for neutrinos, long-lived particle searches, milli-charged particle searches, QCD, dark matter, dark sectors, and cosmic rays. The FPF will be located 500 m from the ATLAS interaction point. It is shielded from the ATLAS interaction point by 100 m of concrete and rock, creating an extremely low-background environment, ideal for many standard model studies and new physics searches. In this Letter of Interest, we describe the FPF’s location and general features, its physics potential in the HL-LHC era, and topics for further study.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2020

Detecting and studying high-energy collider neutrinos with FASER at the LHC

Neutrinos are copiously produced at particle colliders, but no collider neutrino has ever been detected. Colliders produce both neutrinos and anti-neutrinos of all flavors at very high energies, and they are therefore highly complementary to those from other sources. FASER, the Forward Search Experiment at the LHC, is ideally located to provide the first detection and study of collider neutrinos. We investigate the prospects for neutrino studies with FASER$ν$ , a proposed component of FASER, consisting of emulsion films interleaved with tungsten plates with a total target mass of 1.2 t, to be placed on-axis at the front of FASER. We estimate the neutrino fluxes and interaction rates, describe the FASER$ν$ detector, and analyze the characteristics of the signals and primary backgrounds. For an integrated luminosity of 150 fb –1 to be collected during Run 3 of the 14 TeV LHC in 2021-23, approximately 1300 electron neutrinos, 20,000 muon neutrinos, and 20 tau neutrinos will interact in FASER$ν$, with mean energies of 600 GeV to 1 TeV. With such rates and energies, FASER will measure neutrino cross sections at energies where they are currently unconstrained, will bound models of forward particle production, and could open a new window on physics beyond the standard model.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2019

Analyzing honeycomb photonic crystal waveguides by Dirichlet-to-Neumann maps

As an analog to graphene, honeycomb photonic crystals (PhCs) have attracted a great deal of interest in recent years. The additional degrees of freedom in a honeycomb PhC are useful in designing and optimizing waveguides, can be used to control the symmetry of the system and to realize novel optical devices. In this paper, we present an efficient numerical method for analyzing two-dimensional honeycomb PhC waveguides. Our method is based on a special supercell that covers one period of the honeycomb PhC waveguide and the so-called Dirichlet-to-Neumann map of the supercell. The method gives rise to a linear eigenvalue value problem with relatively small matrices, and it is validated by comparing with previous works. As an application of the method, we calculate edge modes for a few different PhCs. For a honeycomb PhC with dielectric rods in air, we found an edge mode at the zigzag edge, without using special modified rods at the edge. For a honeycomb PhC with air holes, we found a new edge mode when the PhC is bounded by a perfectly electric conductor.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗
DOE OSTI2019

A finalized determinant for complete lignocellulose enzymatic saccharification potential to maximize bioethanol production in bioenergy Miscanthus

Miscanthus is a leading bioenergy crop with enormous lignocellulose production potential for biofuels and chemicals. However, lignocellulose recalcitrance leads to biomass process difficulty for an efficient bioethanol production. Hence, it becomes essential to identify the integrative impact of lignocellulose recalcitrant factors on cellulose accessibility for biomass enzymatic hydrolysis. In this study, we analyzed four typical pairs of Miscanthus accessions that showed distinct cell wall compositions and sorted out three major factors that affected biomass saccharification for maximum bioethanol production. Among the three optimal (i.e., liquid hot water, H 2 SO 4 and NaOH) pretreatments performed, mild alkali pretreatment (4% NaOH at 50 °C) led to almost complete biomass saccharification when 1% Tween-80 was co-supplied into enzymatic hydrolysis in the desirable Miscanthus accessions. Consequently, the highest bioethanol yields were obtained at 19% (% dry matter) from yeast fermentation, with much higher sugar–ethanol conversion rates by 94–98%, compared to the other Miscanthus species subjected to stronger pretreatments as reported in previous studies. By comparison, three optimized pretreatments distinctively extracted wall polymers and specifically altered polymer features and inter-linkage styles, but the alkali pretreatment caused much increased biomass porosity than that of the other pretreatments. Based on integrative analyses, excellent equations were generated to precisely estimate hexoses and ethanol yields under various pretreatments and a hypothetical model was proposed to outline an integrative impact on biomass saccharification and bioethanol production subjective to a predominate factor (CR stain) of biomass porosity and four additional minor factors (DY stain, cellulose DP, hemicellulose X/A, lignin G-monomer). Using four pairs of Miscanthus samples with distinct cell wall composition and varied biomass saccharification, this study has determined three main factors of lignocellulose recalcitrance that could be significantly reduced for much-increased biomass porosity upon optimal pretreatments. It has also established a novel standard that should be applicable to judge any types of biomass process technology for high biofuel production in distinct lignocellulose substrates. Hence, this study provides a potential strategy for precise genetic modification of lignocellulose in all bioenergy crops.

09 BIOMASS FUELS↗
DOE OSTI2019

Phosphate Sorption Speciation and Precipitation Mechanisms on Amorphous Aluminum Hydroxide

Aluminum (Al) oxides are important adsorbents for phosphate in soils and sediments, and significantly limit Phosphate (P) mobility and bioavailability, but the speciation of surface-adsorbed phosphate on Al oxides remains poorly understood. Here, phosphate sorption speciation on amorphous Al hydroxide (AAH) was determined under pH 3–8 and P concentration of 0.03 mM–15 mM using various spectroscopic approaches, and phosphate precipitation mechanisms were discussed as well. AAH exhibits an extremely high phosphate sorption capacity, increasing from 3.80 mmol/g at pH 7 to 4.63 mmol/g at pH 3. Regardless of reaction pH, with increasing P sorption loading, the sorption mechanism transits from bidentate binuclear (BB) surface complexation with dP-Al of 3.12 Å to surface precipitation of analogous amorphous AlPO4 (AAP), possibly with ternary complexes, such as (≡Al-O) 2 -PO 2 -Al, as intermediate products. Additionally, the percentage of precipitated phosphate occurring in AAP linearly and positively correlates with P sorption loading. Compared to phosphate reaction with ferrihydrite, phosphate adsorbs and precipitates more readily on AAH due to the higher solubility product (K sp ) of AAH. The formation of AAP particles involves Al III release, which is promoted by phosphate adsorption, and its subsequent precipitation with phosphate at AAH surfaces or in the bulk solution.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2018

An FPGA-based Pattern Recognition Associative Memory

Pattern recognition associative memory (PRAM) devices are parallel processing engines which are used to tackle the complex combinatorics of track nding algorithms, particularly for silicon based tracking triggers. PRAM development has been mostly limited to the realm of ASICs, which often leads to lengthy and expensive design cycles. FPGAs allow for quick iterations, making them an ideal hardware platform for designing and evaluating new PRAM features before committing to silicon. In this paper we present our FPGA-based PRAM designs and discuss how logic blocks which were originally developed for ASICs have been redesigned to take advantage of modern FPGA architectures.

43 PARTICLE ACCELERATORS↗
DOE OSTI2018

Genetic loci simultaneously controlling lignin monomers and biomass digestibility of rice straw

Lignin content and composition are crucial factors affecting biomass digestibility. Exploring the genetic loci simultaneously affecting lignin-relevant traits and biomass digestibility is a precondition for lignin genetic manipulation towards energy crop breeding. In this study, a high-throughput platform was employed to assay the lignin content, lignin composition and biomass enzymatic digestibility of a rice recombinant inbred line population. Correlation analysis indicated that the absolute content of lignin monomers rather than lignin content had negative effects on biomass saccharification, whereas the relative content of p-hydroxyphenyl unit and the molar ratio of p-hydroxyphenyl unit to guaiacyl unit exhibited positive roles. Eight QTL clusters were identified and four of them affecting both lignin composition and biomass digestibility. Here, the additive effects of clustered QTL revealed consistent relationships between lignin-relevant traits and biomass digestibility. Pyramiding rice lines containing the above four positive alleles for increasing biomass digestibility were selected and showed comparable lignin content, decreased syringyl or guaiacyl unit and increased molar percentage of p-hydroxyphenyl unit, the molar ratio of p-hydroxyphenyl unit to guaiacyl unit and sugar releases. More importantly, the lodging resistance and eating/cooking quality of pyramiding lines were not sacrificed, indicating the QTL information could be applied to select desirable energy rice lines.

59 BASIC BIOLOGICAL SCIENCES↗