Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Time Series”

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.

At least 73 records · Page 4

A Preferences Corpus and Annotation Scheme for Human-Guided Alignment of Time-Series GPTs

The process of time-series forecasting such as predicting trajectories of silicon content in blast furnaces is a difficult task. Most time-series approaches today focus on scalar-type MSE loss optimization. This optimization approach, while widely common, could benefit from the use of human expert or process-level preferences. In this paper, we introduce a novel alignment and fine-tuning approach that involves learning from a corpus of preferred and dis-preferred time-series prediction trajectories. Our contributions include (1) a preference annotation pipeline for time-series forecasts, (2) the application of Score-based Preference Optimization (SPO) to train decoder-only transformers from preferences, and (3) results showing improvements in forecast quality. The approach is validated on both proprietary blast furnace data and the UCI Appliances Energy dataset. The proposed preference corpus and training strategy offer a new option for fine-tuning sequence models in industrial settings.

DPO↗

Array magnetics modal analysis for the DIII-D tokamak based on localized time-series modelling

Here, time-series analysis of magnetics data in tokamaks is typically done using block-based fast Fourier transform methods. This work presents the development and deployment of a new set of algorithms for magnetic probe array analysis. The method is based on an estimation technique known as stochastic subspace identification (SSI). Compared with the standard coherence approach or the direct singular value decomposition approach, the new technique exhibits several beneficial properties. For example, the SSI method does not require that frequencies are orthogonal with respect to the timeframe used in the analysis. Frequencies are obtained directly as parameters of localized time-series models. The parameters are extracted by solving small-scale eigenvalue problems. Applications include maximum-likelihood regularized eigenmode pattern estimation, detection of neoclassical tearing modes, including locked mode precursors, and automatic clustering of modes, and magnetics-pattern characterization of sawtooth pre- and postcursors, edge harmonic oscillations and fishbones.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

A Waveform Skewness Index for Measuring Time Series Nonlinearity and Its Applications to the ENSO-Indian Monsoon Relationship

Many geophysical time series possess nonlinear characteristics that reflect the underlying physics of the phenomena the time series describe. The nonlinear character of times series can change with time, so it is important to quantify time series nonlinearity without assuming stationarity. A common way of quantifying the time evolution of time series nonlinearity is to compute sliding skewness time series, but it is shown here that such an approach can be misleading when time series contain periodicities. To remedy this deficiency of skewness, a new waveform skewness index is proposed for quantifying local nonlinearities embedded in time series. A waveform skewness spectrum is proposed for determining the frequency components that are contributing to time series waveform skewness. The new methods are applied to the El Niño–Southern Oscillation (ENSO) and the Indian monsoon to test a recently proposed hypothesis that states that changes in the ENSO–Indian monsoon relationship are related to ENSO nonlinearity. We show that the ENSO–Indian rainfall relationship weakens during time periods of high ENSO waveform skewness. The results from two different analyses suggest that the breakdown of the ENSO–Indian monsoon relationship during time periods of high ENSO waveform skewness is related to the more frequent occurrence of strong central Pacific El Niño events, supporting arguments that changes in the ENSO–Indian rainfall relationship are not solely related to noise.

Justin Schulte↗

Statistical criteria for characterizing irradiance time series.

We propose and examine several statistical criteria for characterizing time series of solar irradiance. Time series of irradiance are used in analyses that seek to quantify the performance of photovoltaic (PV) power systems over time. Time series of irradiance are either measured or are simulated using models. Simulations of irradiance are often calibrated to or generated from statistics for observed irradiance and simulations are validated by comparing the simulation output to the observed irradiance. Criteria used in this comparison should derive from the context of the analyses in which the simulated irradiance is to be used. We examine three statistics that characterize time series and their use as criteria for comparing time series. We demonstrate these statistics using observed irradiance data recorded in August 2007 in Las Vegas, Nevada, and in June 2009 in Albuquerque, New Mexico.

24 POWER TRANSMISSION AND DISTRIBUTION↗

Time Series Analysis of Insar Data: Methods and Trends

Time series analysis of InSAR data has emerged as an important tool for monitoring and measuring the displacement of the Earth's surface. Changes in the Earth's surface can result from a wide range of phenomena such as earthquakes, volcanoes, landslides, variations in ground water levels, and changes in wetland water levels. Time series analysis is applied to interferometric phase measurements, which wrap around when the observed motion is larger than one-half of the radar wavelength. Thus, the spatio-temporal ''unwrapping" of phase observations is necessary to obtain physically meaningful results. Several different algorithms have been developed for time series analysis of InSAR data to solve for this ambiguity. These algorithms may employ different models for time series analysis, but they all generate a first-order deformation rate, which can be compared to each other. However, there is no single algorithm that can provide optimal results in all cases. Since time series analyses of InSAR data are used in a variety of applications with different characteristics, each algorithm possesses inherently unique strengths and weaknesses. In this review article, following a brief overview of InSAR technology, we discuss several algorithms developed for time series analysis of InSAR data using an example set of results for measuring subsidence rates in Mexico City.

InSAR↗

Energy Efficient Streaming Time Series Classification with Attentive Power Iteration

Efficiently processing time series data streams in real-time on resource-constrained devices offers significant advantages in terms of enhanced computational energy efficiency and reduced time-related risks. We introduce an innovative streaming time series classification network that utilizes attentive power iteration, enabling real-time processing on resource-constrained devices. Our model continuously updates a compact representation of the entire time series, enhancing classification accuracy while conserving energy and processing time. Notably, it excels in streaming scenarios without requiring complete time series access, enabling swift decisions. Experimental results show that our approach excels in classification accuracy and energy efficiency, with over 70% less consumption and threefold faster task completion than benchmarks. This work advances real-time responsiveness, energy conservation, and operational effectiveness for constrained devices, contributing to optimizing various applications.

97 MATHEMATICS AND COMPUTING↗

SigTime: Learning and Visually Explaining Time Series Signatures

Understanding and distinguishing temporal patterns in time series data is essential for scientific discovery and decision-making. For example, in biomedical research, uncovering meaningful patterns in physiological signals can improve diagnosis, risk assessment, and patient outcomes. However, existing methods for time series pattern discovery face major challenges, including high computational complexity, limited interpretability, and difficulty in capturing meaningful temporal structures. Here, to address these gaps, we introduce a novel learning framework that jointly trains two Transformer models using complementary time series representations: shapelet-based representations to capture localized temporal structures and traditional feature engineering to encode statistical properties. The learned shapelets serve as interpretable signatures that differentiate time series across classification labels. Additionally, we develop a visual analytics system—SigTime—with coordinated views to facilitate exploration of time series signatures from multiple perspectives, aiding in useful insights generation. We quantitatively evaluate our learning framework on eight publicly available datasets and one proprietary clinical dataset. Additionally, we demonstrate the effectiveness of our system through two usage scenarios along with the domain experts: one involving public ECG data and the other focused on preterm labor analysis.

97 MATHEMATICS AND COMPUTING↗

Clustering Analysis of Commercial Vehicles Using Automatically Extracted Features from Time Series Data

Standard of practice approaches to time series cluster analysis involve careful feature engineering, often utilizing expert input to tune and select features by hand. In many cases, expert input may not be readily available, or there may not yet exist a community consensus on the ideal features for a given application. This paper compares the results of several cluster analysis methods, using both hand selected features and those extracted automatically, when applied to large geospatial time series telematics data from commercial trucking fleets. The impacts of feature selection, dimensionality reduction, and choice of clustering algorithm on the quality of clustering results are explored. Results from this analysis confirm prior results that domain agnostic features are competitive with the hand engineered features with respect to clustering quality metrics. These results also provide new insight into the most successful strategies for identifying structure in large unstructured vehicle telematics data, and suggest that time series clustering using automatic feature extraction can be an effective approach to extract structure from large scale geospatial time series data in cases when hand selected features are not available.

33 ADVANCED PROPULSION SYSTEMS↗

Deep learning for time series forecasting: a survey of recent advances

Time series forecasting plays a critical role in numerous real-world applications, such as finance, healthcare, transportation, and scientific computing. In recent years, deep learning has become a powerful tool for modeling complex temporal patterns and improving forecasting accuracy. This survey provides an overview of recent deep learning approaches for time series forecasting, involving various architectures including RNNs, CNNs, GNNs, transformers, large language models, MLP-based models, and diffusion models. We first identify key challenges in the field, such as temporal dependency, efficiency, and cross-variable dependency, which drive the development of forecasting techniques. Then, the general advantages and limitations of each architecture are discussed to contextualize their adaptation in time series forecasting. Furthermore, we highlight promising design trends like multi-scale modeling, decomposition, and frequency-domain techniques, which are shaping the future of the field. This paper serves as a compact reference for researchers and practitioners seeking to understand the current landscape and future trajectory of deep learning in time series forecasting.

97 MATHEMATICS AND COMPUTING↗

Simulation of Ground Winds Time Series for the NASA Crew Launch Vehicle (CLV)

Simulation of wind time series based on power spectrum density (PSD) and spectral coherence models for ground wind turbulence is described. The wind models, originally developed for the Shuttle program, are based on wind measurements at the NASA 150-m meteorological tower at Cape Canaveral, FL. The current application is for the design and/or protection of the CLV from wind effects during on-pad exposure during periods from as long as days prior to launch, to seconds or minutes just prior to launch and seconds after launch. The evaluation of vehicle response to wind will influence the design and operation of constraint systems for support of the on-pad vehicle. Longitudinal and lateral wind component time series are simulated at critical vehicle locations. The PSD model for wind turbulence is a function of mean wind speed, elevation and temporal frequency. Integration of the PSD equation over a selected frequency range yields the variance of the time series to be simulated. The square root of the PSD defines a low-pass filter that is applied to adjust the components of the Fast Fourier Transform (FFT) of Gaussian white noise. The first simulated time series near the top of the launch vehicle is the inverse transform of the adjusted FFT. Simulation of the wind component time series at the nearest adjacent location (and all other succeeding next nearest locations) is based on a model for the coherence between winds at two locations as a function of frequency and separation distance, where the adjacent locations are separated vertically and/or horizontally. The coherence function is used to calculate a coherence weighted FFT of the wind at the next nearest location, given the FFT of the simulated time series at the previous location and the essentially incoherent FFT of the wind at the selected location derived a priori from the PSD model. The simulated time series at each adjacent location is the inverse Fourier transform of the coherence weighted FFT. For a selected design case, the equations, the process and the simulated time series at multiple vehicle stations are presented.

Adelfang, Stanley I.↗

Probabilistic estimation of depth-resolved profiles of soil thermal diffusivity from temperature time series: Supporting Data

This dataset consists of soil temperature time series that were used to estimate soil thermal diffusivity and its uncertainty trough the probabilistic modelling approach developed and presented in the article named "Probabilistic estimation of depth-resolved profiles of soil thermal diffusivity from temperature time series" and published in Earth Surface Dynamics. There are two compressed (.zip) files that contains synthetic (Synthetic_soiltemp_Teller.zip) and field (Field_soiltemp_Teller.zip) data. There is one MATLAB file that requires MATLAB to execute but any text editor can open it. The Synthetic_soiltemp_Teller.zip file includes 5 comma-delimited data files (.csv) each of which contains soil temperature time series generated through forward modeling (i.e., heat-conduction process in a heterogeneous medium using an explicit finite difference method) to mimic various types of temperature gradients, trend and fluctuations. These synthetic soil temperatures were used to investigate the impact of different environmental conditions on the uncertainty of thermal diffusivity estimates. The Field_soiltemp_Teller.zip file contains 28 comma-delimited data files (.csv) out of which (a) 27 files includes soil temperature time series recorded from 27 temperature probes located in a site along Teller Road about 40 km northwest of Nome, Alaska (64.72°N, 165.94°W), (b) one includes the name and coordinates of the 27 probes. These field soil temperatures were used to infer soil thermal diffusivity at numerous locations and depths in a discontinuous permafrost environment, and to evaluate the links between the estimated soil thermal diffusivity values and soil physical properties. The comma-delimited data files (.csv) of the synthetic and field soil temperature time series includes date and time (UTC) in the first column and soil temperature from 5 cm below the ground surface to 1.05 m depth (with 5 or 10 cm spacing between sensors) in the other columns. The measurements were acquired every 15 minutes. The Next-Generation Ecosystem Experiments: Arctic (NGEE Arctic), was a research effort to reduce uncertainty in Earth System Models by developing a predictive understanding of carbon-rich Arctic ecosystems and feedbacks to climate. NGEE Arctic was supported by the Department of Energy's Office of Biological and Environmental Research. The NGEE Arctic project had two field research sites: 1) located within the Arctic polygonal tundra coastal region on the Barrow Environmental Observatory (BEO) and the North Slope near Utqiagvik (Barrow), Alaska and 2) multiple areas on the discontinuous permafrost region of the Seward Peninsula north of Nome, Alaska. Through observations, experiments, and synthesis with existing datasets, NGEE Arctic provided an enhanced knowledge base for multi-scale modeling and contributed to improved process representation at global pan-Arctic scales within the Department of Energy's Earth system Model (the Energy Exascale Earth System Model, or E3SM), and specifically within the E3SM Land Model component (ELM).

54 ENVIRONMENTAL SCIENCES↗

GrammarViz 3.0: Interactive Discovery of Variable-Length Time Series Patterns

The problems of recurrent and anomalous pattern discovery in time series, e.g., motifs and discords, respectively, have received a lot of attention from researchers in the past decade. However, since the pattern search space is usually intractable, most existing detection algorithms require that the patterns have discriminative characteristics and have its length known in advance and provided as input, which is an unreasonable requirement for many real-world problems. In addition, patterns of similar structure, but of different lengths may co-exist in a time series. In order to address these issues, we have developed algorithms for variable-length time series pattern discovery that are based on symbolic discretization and grammar inference—two techniques whose combination enables the structured reduction of the search space and discovery of the candidate patterns in linear time. In this work, we present GrammarViz 3.0—a software package that provides implementations of proposed algorithms and graphical user interface for interactive variable-length time series pattern discovery. The current version of the software provides an alternative grammar inference algorithm that improves the time series motif discovery workflow, and introduces an experimental procedure for automated discretization parameter selection that builds upon the minimum cardinality maximum cover principle and aids the time series recurrent and anomalous pattern discovery.

97 MATHEMATICS AND COMPUTING↗

Assessing methods in fusion and fitting for time series construction in remote sensing-based earth observations

This study evaluates the comparative performance of spatiotemporal fusion and time-series fitting methods for constructing high-spatiotemporal-resolution remote sensing time-series data. Due to in-class similarity of fusion methods and fitting methods, we employ the Fit-FC (Fitting, spatial Filtering, and residual Compensation) model as a representative fusion method and the linear harmonic fitting model as a representative fitting method. Both Fit-FC and the linear harmonic fitting are widely used for high-spatiotemporal-resolution time-series data construction, and we modify the original Fit-FC model to enable automatic time-series fusion. To ensure data representativeness, we use 3 years (2019–2021) of Harmonized Landsat and Sentinel-2 surface reflectance datasets and Terra MCD43A4 products. Eight experimental regions are selected worldwide to guarantee generalization of the comparative performance between fusion and fitting methods, covering diverse land-use types (cropland, developed land, forest, and grassland) and varying climatological conditions. Time-series of NDVI and surface reflectance are analyzed under both actual observations and simulated data-missing scenarios. The constructed time-series data reveals that (1) the modified Fit-FC and linear harmonic fitting model achieve excellent performance in constructing high-resolution time-series images; (2) the fusion method outperforms the fitting method in constructing time-series of NDVI and surface reflectance images in cropland-, forest-, and grassland-dominated regions; (3) both methods achieve comparable performance in developed-dominated regions; (4) the fusion method is more robust to missing data, and better captures abrupt phenological transitions under conditions of continuous missing data; (5) the fitting method is computationally more efficient, making it suitable for large-scale time-series image reconstruction. This study provides valuable insights for selecting optimal strategies to generate high-resolution time-series images across diverse application scenarios and lays a foundation for extensions to other vegetation indices or land surface variables.

54 ENVIRONMENTAL SCIENCES↗

Deep Cellular Recurrent Network for Efficient Analysis of Time-Series Data With Spatial Information

Efficient processing of large-scale time series data is an intricate problem in machine learning. Conventional sensor signal processing pipelines with hand engineered feature extraction often involve huge computational cost with high dimensional data. Deep recurrent neural networks have shown promise in automated feature learning for improved time-series processing. However, generic deep recurrent models grow in scale and depth with increased complexity of the data. This is particularly challenging in presence of high dimensional data with temporal and spatial characteristics. Consequently, this work proposes a novel deep cellular recurrent neural network (DCRNN) architecture to efficiently process complex multi-dimensional time series data with spatial information. Here, the cellular recurrent architecture in the proposed model allows for location-aware synchronous processing of time series data from spatially distributed sensor signal sources. Extensive trainable parameter sharing due to cellularity in the proposed architecture ensures efficiency in the use of recurrent processing units with high-dimensional inputs. This study also investigates the versatility of the proposed DCRNN model for classification of multi-class time series data from different application domains. Consequently, the proposed DCRNN architecture is evaluated using two time-series datasets: a multichannel scalp EEG dataset for seizure detection, and a machine fault detection dataset obtained in-house. The results suggest that the proposed architecture achieves state-of-the-art performance while utilizing substantially less trainable parameters when compared to comparable methods in the literature.

60 APPLIED LIFE SCIENCES↗

Fast and Flexible Multivariate Time Series Subsequence Search

Multivariate Time-Series (MTS) are ubiquitous, and are generated in areas as disparate as sensor recordings in aerospace systems, music and video streams, medical monitoring, and financial systems. Domain experts are often interested in searching for interesting multivariate patterns from these MTS databases which often contain several gigabytes of data. Surprisingly, research on MTS search is very limited. Most of the existing work only supports queries with the same length of data, or queries on a fixed set of variables. In this paper, we propose an efficient and flexible subsequence search framework for massive MTS databases, that, for the first time, enables querying on any subset of variables with arbitrary time delays between them. We propose two algorithms to solve this problem (1) a List Based Search (LBS) algorithm which uses sorted lists for indexing, and (2) a R*-tree Based Search (RBS) which uses Minimum Bounding Rectangles (MBR) to organize the subsequences. Both algorithms guarantee that all matching patterns within the specified thresholds will be returned (no false dismissals). The very few false alarms can be removed by a post-processing step. Since our framework is also capable of Univariate Time-Series (UTS) subsequence search, we first demonstrate the efficiency of our algorithms on several UTS datasets previously used in the literature. We follow this up with experiments using two large MTS databases from the aviation domain, each containing several millions of observations. Both these tests show that our algorithms have very high prune rates (>99%) thus needing actual disk access for only less than 1% of the observations. To the best of our knowledge, MTS subsequence search has never been attempted on datasets of the size we have used in this paper.

Bhaduri, Kanishka↗

Using NASA's Giovanni System to Simulate Time-Series Stations in the Outflow Region of California's Eel River

Oceanographic time-series stations provide vital data for the monitoring of oceanic processes, particularly those associated with trends over time and interannual variability. There are likely numerous locations where the establishment of a time-series station would be desirable, but for reasons of funding or logistics, such establishment may not be feasible. An alternative to an operational time-series station is monitoring of sites via remote sensing. In this study, the NASA Giovanni data system is employed to simulate the establishment of two time-series stations near the outflow region of California s Eel River, which carries a high sediment load. Previous time-series analysis of this location (Acker et al. 2009) indicated that remotely-sensed chl a exhibits a statistically significant increasing trend during summer (low flow) months, but no apparent trend during winter (high flow) months. Examination of several newly-available ocean data parameters in Giovanni, including 8-day resolution data, demonstrates the differences in ocean parameter trends at the two locations compared to regionally-averaged time-series. The hypothesis that the increased summer chl a values are related to increasing SST is evaluated, and the signature of the Eel River plume is defined with ocean optical parameters.

Acker, James G.↗

Bayesian classification of partially observed outbreaks using time-series data.

Results show that a time-series based classification may be possible. For the test cases considered, the correct model can be selected and the number of index case can be captured within {+-} {sigma} with 5-10 days of data. The low signal-to-noise ratio makes the classification difficult for small epidemics. The problem statement is: (1) Create Bayesian techniques to classify and characterize epidemics from a time-series of ICD-9 codes (will call this time-series a 'morbidity stream'); and (2) It is assumed the morbidity stream has already set off an alarm (through a Kalman filter anomaly detector) Starting with a set of putative diseases: Identify which disease or set of diseases 'fit the data best' and, Infer associated information about it, i.e. number of index cases, start time of the epidemic, spread rate, etc.

99 GENERAL AND MISCELLANEOUS↗

Early Time-Series Classification with Reliability Guarantee

We consider the early classification of (incomplete) time-series data given a complete time series training set. The early classification problem arises naturally when test sample data is collected over time, or when costs must be incurred to collect the data. For example, for missile defense, it is important to determine the target type long before it reaches its target. A practical goal is to assign a class label as soon as enough data is available to make a good decision. This objective is formalized through the notion of reliability—the probability that a label assigned to the early, incomplete data matches that assigned to the complete data, and we propose a method to classify incomplete data only if a user-specified reliability threshold is met. Our approach models the complete data as a random variable whose distribution is dependent on the current incomplete data and the training data. The method differs from standard strategies in that our focus is on determining the reliability of the early classification decision, not only the accuracy. Proposed methods are tested on a set of open-domain time-series datasets; where the goal is to classify the time-series as early as possible while still guaranteeing that the reliability threshold is met.

Anderson, Hyrum S.↗