Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “ARIMA”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

11 records
All sourcesNASA NTRSDOE OSTI
DOE OSTI2025

Temporal Forecasting of Distributed Temperature Sensing in a Thermal Hydraulic System With Machine Learning and Statistical Models

We benchmark performance of long-short term memory (LSTM) network machine learning model and autoregressive integrated moving average (ARIMA) statistical model in temporal forecasting of distributed temperature sensing (DTS). Data in this study consists of fluid temperature transient measured with two co-located Rayleigh scattering fiber optic sensors (FOS) in a forced convection mixing zone of a thermal tee. We treat each gauge of a FOS as an independent temperature sensor. We first study prediction of DTS time series using Vanilla LSTM and ARIMA models trained on prior history of the same FOS that is used for testing. The results yield maximum absolute percentage error (MaxAPE) and root mean squared percentage error (RMSPE) of 1.58% and 0.06% for ARIMA, and 3.14% and 0.44% for LSTM, respectively. Next, we investigate zero-shot forecasting (ZSF) with LSTM and ARIMA trained on history of the co-located FOS only, which is advantageous when limited training data is available. The ZSF MaxAPE and RMSPE values for ARIMA are comparable to those of the Vanilla use case, while the error values for LSTM increase. We show that in ZSF, performance of LSTM network can be improved by training on most correlated gauges between the two FOS, which are identified by calculating the Pearson correlation coefficient. The improved ZSF MaxAPE and RMSPE for LSTM are 4.4% and 0.33%, respectively. Performance of ZSF LSTM can be further enhanced through transfer learning (TL), where LSTM is re-trained on a subset of the FOS that is the target of forecasting. We show that LSTM pre-trained on correlated dataset and re-trained on 30% of testing target dataset achieves MaxAPE and RMSPE values of 2.32% and 0.28%, respectively.

ARIMA↗
DOE OSTI2024

Comparison of Machine Learning-Based Predictive Models of the Nutrient Loads Delivered from the Mississippi/Atchafalaya River Basin to the Gulf of Mexico

Predicting nutrient loads is essential to understanding and managing one of the environmental issues faced by the northern Gulf of Mexico hypoxic zone, which poses a severe threat to the Gulf’s healthy ecosystem and economy. The development of hypoxia in the Gulf of Mexico is strongly associated with the eutrophication process initiated by excessive nutrient loads. Due to the complexities in the excessive nutrient loads to the Gulf of Mexico, it is challenging to understand and predict the underlying temporal variation of nutrient loads. The study was aimed at identifying an optimal predictive machine learning model to capture and predict nonlinear behavior of the nutrient loads delivered from the Mississippi/Atchafalaya River Basin (MARB) to the Gulf of Mexico. For this purpose, monthly nutrient loads (N and P) in tons were collected from US Geological Survey (USGS) monitoring station 07373420 from 1980 to 2020. Machine learning models—including autoregressive integrated moving average (ARIMA), gaussian process regression (GPR), single-layer multilayer perceptron (MLP), and a long short-term memory (LSTM) with the single hidden layer—were developed to predict the monthly nutrient loads, and model performances were evaluated by standard assessment metrics—Root Mean Square Error (RMSE) and Correlation Coefficient (R). The residuals of predictive models were examined by the Durbin–Watson statistic. The results showed that MLP and LSTM persistently achieved better accuracy in predicting monthly TN and TP loads compared to GPR and ARIMA. In addition, GPR models achieved slightly better test RMSE score than ARIMA models while their correlation coefficients are much lower than ARIMA models. Moreover, MLP performed slightly better than LSTM in predicting monthly TP loads while LSTM slightly outperformed for TN loads. Furthermore, it was found that the optimizer and number of inputs didn’t show effects on the LSTM performance while they exhibited impacts on MLP outcomes. This study explores the capability of machine learning models to accurately predict nonlinearly fluctuating nutrient loads delivered to the Gulf of Mexico. Further efforts focus on improving the accuracy of forecasting using hybrid models which combine several machine learning models with superior predictive performance for nutrient fluxes throughout the MARB.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2026

Reliable statistics-based detection and investigation of anomalies in a SMART valve system

Reliable anomaly detection and diagnosis are critical for the safe operation of complex engineered systems. This study presents a unified framework that integrates statistical, model-based, and data-driven techniques for anomaly detection and investigation, demonstrated on SMART valve systems in hybrid energy applications. Four detection methods—mean deviation, seasonal extreme studentized deviate, ARIMA forecasting, and matrix profiling—were implemented and compared. Matrix profiling was particularly effective in revealing subtle deviations and hidden relationships among variables. Anomaly investigation was performed by analyzing variable-level and grouped signal profiles, with system topology incorporated to distinguish primary faults from propagated effects. Grouping signals by type enhanced interpretability, enabling accurate localization of anomalies across multi-dimensional datasets. Experimental results confirmed the framework's capability to consistently detect and isolate anomalies while providing actionable insights into system interdependencies. The proposed methodology offers a robust, interpretable, and scalable solution for condition monitoring, with potential applications in safety-critical domains such as nuclear energy, aerospace, and process industries.

ARIMA models↗
DOE OSTI2026

A Centralized AI Lakehouse Framework for Brain Tumor MRI Classification and Segmentation, University KPI Forecasting, and Water Potability Prediction

In many university and healthcare projects, models are built for very different data types such as tables, institutional time series, and medical images, but they are deployed as separate applications. In this work, that separation made testing and maintenance difficult because each module had its own pipeline and runtime requirements. This paper presents an integrated AI lakehouse-style implementation that runs three model pipelines inside one containerized backend. For medical imaging, we used MRI datasets from IEEE DataPort: a four-class classification set with 7012 images (5708 train/1304 test) and a segmentation set with 3063 image–mask pairs. The classification model (ResNet50 transfer learning) is evaluated using a proper train–validation–test protocol across multiple splits (80/10/10, 70/10/20, 60/10/30, and 10/30/60), achieving a test accuracy of 99.00% under the standard 80/10/10 split. Additionally, a patient-level evaluation is conducted using an external glioma dataset to provide a more realistic assessment without data leakage. The segmentation model (DeepLabV3-ResNet50) achieved 83.09% validation mIoU and 88.79% Dice score. For university KPI forecasting, we used annual IPEDS and NSF HERD data from 2010 to 2023 for three universities (BSU, EOU, and UAB). To examine the effect of preprocessing on forecasting performance, two case studies are conducted. In the first case, linear interpolation is applied to generate semester-level data. In the second case, the original annual data is used directly without interpolation. Random Forest regression and ARIMA models are evaluated using MAE, RMSE, MAPE, and R 2 . The results showed that interpolation improved apparent forecasting performance due to smoothing, while evaluation on the original annual data provided a more realistic assessment of model behavior. To further validate the framework on a larger dataset, an additional case study is conducted using a student dropout dataset. For water potability, we trained and compared multiple tabular classifiers on a large dataset (1,048,575 samples). A Random Forest model (100 trees, max depth 10) achieved 85.86% test accuracy and high recall for unsafe samples (0.8447). All modules are served via FastAPI and deployed together using Docker, with workflow automation routing requests to the correct endpoint. System-level benchmarking indicates that the backend maintains stable throughput and latency under concurrent requests.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2026

A comprehensive framework to assess elemental mercury in the Department of Energy: A time series analysis

Objective: This study investigated whether seasonal categories affect airborne mercury concentrations in the U.S. Department of Energy operations. Methods: We conducted an initial assessment of the general variability of airborne elemental mercury time-weighted average (TWA) samples. Then, we performed a two-component time series analysis to determine whether long-term, cyclical temperature change patterns affect mercury concentrations. Results: Both ARIMA time series models demonstrated stationary, non-random means (χ² = 83.8, p < 0.001) and standard deviation (χ² = 55.8, p < 0.001) of mercury concentrations. Here, our results indicate that the seasonal factors did not influence mercury concentration. Conclusions: Our results demonstrate that mercury concentrations primarily emanate from operational activities, work practices, and/or transient environmental conditions rather than seasonal fluctuations.

Cannady, Ryan T. [Oak Ridge National Laboratory (O↗
DOE OSTI2024

Evaluating the Accuracy of Machine Learning Forecasts

To improve the accuracy of forecasting in machine learning, we must investigate multiple machine learning models and see how accurately they can predict values after training. We used seven machine learning models to try and get more accurate predictions. The models that were used were ARIMA, SES, MLP, CART, LightGBM, and XGBoost. We used a processed dataset from a Terminal at LAX that had the number of people traveling through terminal X every hour in March from 2015-2019. We trained our models with the dates March 6 - March 19 to predict the value for March 20th and the hours 6:00 am to 6:00 pm since those are the most popular traveling hours. By using the different models, we had varying results of accuracy when estimating the amount of people traveling through terminal X on March 20th. We know that machine learning models are helpful for forecasting and by seeing how accurately these models can predict, we can see how forecasting can be helpful for other issues. Using these methods, airports can use forecasting to predict the amount of people coming in and out and can use these predictions to prepare their resource management, operational efficiency, and overall passenger experience.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2024

Integrative analysis of the 3D genome and epigenome in mouse embryonic tissues

While a rich set of putative cis-regulatory sequences involved in mouse fetal development have been annotated recently on the basis of chromatin accessibility and histone modification patterns, delineating their role in developmentally regulated gene expression continues to be challenging. To fill this gap, here we mapped chromatin contacts between gene promoters and distal sequences across the genome in seven mouse fetal tissues and across six developmental stages of the forebrain. We identified 248,620 long-range chromatin interactions centered at 14,138 protein-coding genes and characterized their tissue-to-tissue variations and developmental dynamics. Integrative analysis of the interactome with previous epigenome and transcriptome datasets from the same tissues revealed a strong correlation between the chromatin contacts and chromatin state at distal enhancers, as well as gene expression patterns at predicted target genes. We predicted target genes of 15,098 candidate enhancers and used them to annotate target genes of homologous candidate enhancers in the human genome that harbor risk variants of human diseases. We present evidence that schizophrenia and other adult disease risk variants are frequently found in fetal enhancers, providing support for the hypothesis of fetal origins of adult diseases.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2025

Measurement of electrical characteristics of Ge-based diodes with thermal GeO2 films exposed to controlled humidity conditions

We investigated the impact of the interaction between gas-phase water molecules and thermally oxidized GeO2 on Ge on the performance of metal–oxide–semiconductor (MOS) structures. A vacuum-integrated setup was developed to form MOS diodes with a GeO2/Ge structure that had been exposed to controlled humidity conditions and to subsequently measure their electrical characteristics in situ. The capacitance–voltage (C–V) curves exhibited a significant negative shift, indicating the generation of positive charges at the GeO2/Ge interface when the GeO2 surface was exposed to humidity levels above approximately 1%. According to a previous study using electron spectroscopy, this threshold corresponds to the humidity level at which molecular water begins to grow on a GeO2/Ge structure. It is likely that gas-phase water molecules infiltrating the GeO2 film bind to local OH sites via hydrogen bonding, leading to the formation of positive fixed charges at the GeO2/Ge interface.

Sano, Shuto↗
DOE OSTI2024

Phonon excitations in Eu 2 Ir 2 O 7 probed by inelastic x-ray scattering

The study of phonon dynamics and its interplay with magnetic ordering is crucial for understanding the unique quantum phases in the pyrochlore iridates. Here, through inelastic x-ray scattering on a single crystal sample of the pyrochlore iridate Eu 2 Ir 2 O 7 , we map out the phonon excitation spectra in Eu 2 Ir 2 O 7 and compare them with the theoretical phonon spectra calculated using the density functional theory. Further, possible phonon renormalization across the magnetic long-range order transition is observed in our experiments, which is consistent with the results of the previous Raman scattering experiments.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗
DOE OSTI2026

Collinear ferromagnetism with reduced moment length in kagome magnet Nd 3 ⁢Ru 4 ⁢Al 12

Here, we determine the magnetic ground state of the kagome lattice magnet Nd 3 ⁢Ru 4⁢ Al 12 by single-crystal neutron diffraction, supported by experiments with polarized neutrons. We identify this material as a collinear ferromagnet (“hex-FM”) with uniform moment length and ordering vector 𝑸 = 0, in contrast to a previous, seminal report that proposed unequal moment lengths on two Nd sites, here called the “ortho-FM” state. Our analysis of the flipping ratio in polarized neutron scattering is consistent with the hex-FM state. The results provide a microscopic basis for understanding the large fluctuation-induced Hall and Nernst responses near 𝑇 C ≈ 41K, as previously reported for Nd 3 ⁢Ru 4 ⁢Al 12 .

RKKY interaction↗
DOE OSTI2025

Dipolar and quadrupolar correlations in the 5⁢𝑑 2 Re-based double perovskites Ba 2 ⁢ YReO 6 and Ba 2 ⁢ ScReO 6

Double perovskites containing heavy transition-metal ions are an important family of compounds for the study of the interplay between electron correlation and spin-orbit coupling. Here, by combining magnetic susceptibility, heat capacity, and neutron scattering measurements, we investigate the dipolar and quadrupolar correlations in two prototypal rhenium-based double perovskite compounds Ba 2 ⁢YReO 6 and Ba 2 ⁢ ScReO 6 . A type-I dipolar antiferromagnetic ground state with a propagation vector 𝐪 = (0,0,1) is observed in both compounds. At temperatures above the magnetic transitions, a quadrupolar ordered phase is identified. Weak spin excitations, which are gapped at low temperatures and softened in the correlated paramagnetic phase, are explained using a minimal model that considers both the dipolar and quadrupolar interactions. In conclusion, at larger wave vectors, we observe dominant phonon excitations that are well described by density functional calculations.

Omar, Otkur [University of Science and Technology ↗