Search NASA⌕ Search

SEARCH · Search NASA

Results for “Complex Network Data”

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 1,351 records · Page 75

Machine Learning‐Guided Discovery of High‐Entropy Perovskite Oxide Electrocatalysts via Oxygen Vacancy Engineering

Abstract High‐entropy perovskite oxides (HEPOs) have recently emerged as multifunctional catalysts. However, the HEPOs’ structural and compositional complexity hinders the easy and accurate extrapolation of activity indicators, which are essential for establishing structure‐property correlations. Here, OxiGraphX, is introduced as a novel graph neural network (GNN) model designed to capture the complex relationships among structure, composition, and atomic chemical environments for accurate prediction of oxygen vacancy formation energies (OVFEs) in HEPOs. By integrating machine learning (ML), density functional theory (DFT), and experimental validation, this work demonstrates an efficient framework for rapidly and accurately screening HEPO electrocatalysts for oxygen evolution reaction (OER). The OxiGraphX predicts OVFEs with a precision exceeding existing data, enabling the identification of compositions of higher oxygen vacancy content (OVC) and, thus, higher catalytic activity. Furthermore, the model explores latent spaces that translate effectively into experimental domains, bridging computational predictions with real‐world applications. This approach accelerates the discovery of high‐performance HEPO catalysts while providing deeper insights into their catalytic mechanisms.

Chemistry↗

RISE: Reducing I/O Contention in Staging-based Extreme-Scale In-situ Workflows

While in-situ workflow formulations have addressed some of the data-related challenges associated with extreme-scale scientific workflows, these workflows involve complex interactions and different modes of data exchange. In the context of increasing system complexity, such workflows present significant resource management challenges, requiring complex cost-performance tradeoffs. This paper presents RISE, an intelligent staging-based data management middleware, which builds on the DataSpaces framework and performs intelligent scheduling of data management operations to reduce I/O contention. In RISE, data are always written immediately to local buffers to reduce the effect of the transfer impact upon application performance. RISE identifies applications’ data access patterns and moves data towards data consumers only when the network is expected to be idle, reducing the impact of asynchronous background data movement upon critical data read/write requests. Here, we experimentally demonstrate that RISE can take advantage of staging nodes to offload data during writes without degrading application data movement performance.

97 MATHEMATICS AND COMPUTING↗

Model-Based Framework to Optimize Charger Station Deployment for Battery Electric Vehicles

The development of battery electric vehicles (BEVs) is accelerating due to their environmental advantages over gasoline and diesel-powered vehicles, including a decrease in air pollution and an increase in energy efficiency. The deployment of charging infrastructure will need to increase to keep pace with demand, especially for large commercial vehicles for which few public chargers currently exist. In this paper, a new flexible framework is proposed for optimizing the placement of charging stations for BEVs, within which different physical models and optimization techniques may be used. Furthermore, a set of metrics is suggested to help enforce complex constraints and facilitate direct comparison between different optimization techniques. Unlike many existing charger placement techniques, the proposed method directly considers the historical driving patterns on a vehicle-by-vehicle basis, using transparent models to assess impacts of candidate charger placements, thus improving the explainability of the results. In the developed framework, modeled BEVs are first generated along the road network to mimic historical traffic data and are simulated traveling along a given route according to a simplified vehicle model. During the simulation, the charger placement problem is initially relaxed to allow vehicles to charge at any node along the road network, and vehicle states are tracked to assess areas of high charging demand. Charging stations are then placed based on the results of the relaxed simulation, and suggested placements are evaluated via road network simulation with fixed charger locations. This proposed framework is applied to a sample problem of placing charging stations along five major highway corridors for Class 8 over-the-road electric trucks. A novel mixed integer programming (MIP) formulation is proposed to optimize charger placements based upon the expected charging demand. Constraints were imposed on the final placement results to limit expected wait times at each station and ensure a minimum threshold of trucking routes are viable for BEVs. The results demonstrate the flexibility and potential effectiveness of the developed model-based framework for scalable charger station deployment.

29 ENERGY PLANNING, POLICY, AND ECONOMY↗

Evolution of context dependent regulation by expansion of feast/famine regulatory proteins

Expansion of transcription factors is believed to have played a crucial role in evolution of all organisms by enabling them to deal with dynamic environments and colonize new environments. We investigated how the expansion of the Feast/Famine Regulatory Protein (FFRP) or Lrp-like proteins into an eight-member family in Halobacterium salinarum NRC-1 has aided in niche-adaptation of this archaeon to a complex and dynamically changing hypersaline environment. We mapped genome-wide binding locations for all eight FFRPs, investigated their preference for binding different effector molecules, and identified the contexts in which they act by analyzing transcriptional responses across 35 growth conditions that mimic different environmental and nutritional conditions this organism is likely to encounter in the wild. Integrative analysis of these data constructed an FFRP regulatory network with conditionally active states that reveal how interrelated variations in DNA-binding domains, effector-molecule preferences, and binding sites in target gene promoters have tuned the functions of each FFRP to the environments in which they act. We demonstrate how conditional regulation of similar genes by two FFRPs, AsnC (an activator) and VNG1237C (a repressor), have striking environment-specific fitness consequences for oxidative stress management and growth, respectively. This study provides a systems perspective into the evolutionary process by which gene duplication within a transcription factor family contributes to environment-specific adaptation of an organism.

59 BASIC BIOLOGICAL SCIENCES↗

Relationships between lineal fracture intensity and chemical composition in the Marcellus Shale, Appalachian Basin

Within mudrock reservoirs, brittle zones undergo failure during hydraulic stimulation, creating numerous artificial fractures which enable hydrocarbons to be liberated from the reservoir. Natural fractures in mudrock reduce the tensile strength of the host rock, creating planes of weaknesses that are hypothesized to be reactivated during hydraulic stimulation. Combined, brittleness and natural fractures contribute to creating more abundant and complex fracture networks during hydraulic stimulation. Research efforts toward quantifying rock brittleness have resulted in numerous mineral-/compositional-based indices, which are used during petrophysical analysis to predict zones most conducive to hydraulic stimulation. In contrast, investigations on the relationship between chemical composition and core-scale natural fractures are limited. For this study, we collected high-resolution energy-dispersive X-ray fluorescence (XRF) data, calibrated with a wave-dispersive XRF, from a Marcellus Shale core. Additionally, we characterized corescale natural fractures in terms of length, width, in-filling material or lack thereof, and orientation. Following the characterization, we transformed the natural fracture data into a continuous P10 (lineal fracture intensity) curve, expressed as the number of fractures per a one-half foot window. Using these data sets, we investigated the relationship between rock composition and natural fracture intensity. Regression analyses recorded positive relationships between natural fracture intensity and calcium, silicon/aluminum, and total organic carbon (TOC), and negative relationships with silicon and aluminum. Aluminum recorded the strongest (negative) relationship ([Formula: see text]) with natural fracture intensity. To access the degree to which natural fractures can be predicted based on chemical composition, we applied a partial least-squares analysis, a multivariate method, and recorded an [Formula: see text]. Our study illustrates that although numerous factors are responsible for natural fracture genesis, such fractures predictively concentrate in areas of similar chemical composition, largely in zones with low aluminum concentrations.

Geochemistry & Geophysics↗

Evaluating Ground Motion Predictions of USGS 3D Seismic Model of the San Francisco Bay Area with Broadband Seismograms

The USGS (Menlo Park) developed a three-dimensional (3D) geologic and seismic model of the greater San Francisco Bay Area for the purposes of computing earthquake ground motions. This model was used to compute scenario ground motions for the 1906 San Francisco, 1989 Loma Prieta and a suite of Hayward Fault earthquakes. While scenario ground motion calculations are important for investigating the amplitude, duration and variability of motions from large damaging earthquakes, the accuracy of such predictions depends on the accuracy of the 3D model. We evaluated the USGS 3D model of the Bay Area by computing predictions of broadband waveforms for 12 moderate (Mw 4-5) earthquakes and comparing them to the observed waveforms. Calculations were performed using WPP (an elastic finite difference code developed at LLNL) on massively parallel computers. Data were obtained from IRIS for Berkeley Digital Seismic Network (BDSN) and USArray broadband stations. The figure below shows a snapshot of the vertical component displacements for an earthquake near Glen Ellen (Sonoma County). Also shown is the comparison of the observed (blue) and computed (red) three-component seismograms at two stations: BKS (Berkeley) and JRSC (Stanford). Note that the motions at BKS are more complex and longer duration due to basin propagation effects from the San Pablo Bay, however the 3D model predicts this energy quite well. The motions at JRSC are simpler and the 3D model predicts the weaker late arriving energy on the transverse (T) component. This analysis found that the USGS 3D model could predict the amplitude, duration and waveform shapes of moderate earthquake ground motions quite well however we did find that phase delays reveal that shear velocities in the model were too fast. This information was used to revise the model for subsequent ground motion modeling.

58 GEOSCIENCES↗

Knowledge-based operation and management of communications systems

Expert systems techniques are being applied in operation and control of the Defense Communications System (DCS), which has the mission of providing reliable worldwide voice, data and message services for U.S. forces and commands. Thousands of personnel operate DCS facilities, and many of their functions match the classical expert system scenario: complex, skill-intensive environments with a full spectrum of problems in training and retention, cost containment, modernization, and so on. Two of these functions are: (1) fault isolation and restoral of dedicated circuits at Tech Control Centers, and (2) network management for the Defense Switched Network (the modernized dial-up voice system currently replacing AUTOVON). An expert system for the first of these is deployed for evaluation purposes at Andrews Air Force Base, and plans are being made for procurement of operational systems. In the second area, knowledge obtained with a sophisticated simulator is being embedded in an expert system. The background, design and status of both projects are described.

Heggestad, Harold M.↗

Expert systems and advanced automation for space missions operations

Increased complexity of space missions during the 1980s led to the introduction of expert systems and advanced automation techniques in mission operations. This paper describes several technologies in operational use or under development at the National Aeronautics and Space Administration's Goddard Space Flight Center. Several expert systems are described that diagnose faults, analyze spacecraft operations and onboard subsystem performance (in conjunction with neural networks), and perform data quality and data accounting functions. The design of customized user interfaces is discussed, with examples of their application to space missions. Displays, which allow mission operators to see the spacecraft position, orientation, and configuration under a variety of operating conditions, are described. Automated systems for scheduling are discussed, and a testbed that allows tests and demonstrations of the associated architectures, interface protocols, and operations concepts is described. Lessons learned are summarized.

Durrani, Sajjad H.↗

A new technology for manufacturing scheduling derived from space system operations

A new technology for producing finite capacity schedules has been developed in response to complex requirements for operating space systems such as the Space Shuttle, the Space Station, and the Deep Space Network for telecommunications. This technology has proven its effectiveness in manufacturing environments where popular scheduling techniques associated with Materials Resources Planning (MRPII) and with factory simulation are not adequate for shop-floor work planning and control. The technology has three components. The first is a set of data structures that accommodate an extremely general description of a factory's resources, its manufacturing activities, and the constraints imposed by the environment. The second component is a language and set of software utilities that enable a rapid synthesis of functional capabilities. The third component is an algorithmic architecture called the Five Ruleset Model which accommodates the unique needs of each factory. Using the new technology, systems can model activities that generate, consume, and/or obligate resources. This allows work-in-process (WIP) to be generated and used; it permits constraints to be imposed or intermediate as well as finished goods inventories. It is also possible to match as closely as possible both the current factory state and future conditions such as promise dates. Schedule revisions can be accommodated without impacting the entire production schedule. Applications have been successful in both discrete and process manufacturing environments. The availability of a high-quality finite capacity production planning capability enhances the data management capabilities of MRP II systems. These schedules can be integrated with shop-floor data collection systems and accounting systems. Using the new technology, semi-custom systems can be developed at costs that are comparable to products that do not have equivalent functional capabilities and/or extensibility.

Hornstein, R. S.↗

Exploring Lightning and Convective Processes Using the Ground-Radar Multiplatform Precipitation Feature Database

The Multiplatform Precipitation Feature (MPF) database synthesizes coincident spaceborne and ground-based lightning and radar data in a framework of storm-based features, fusing broader spaceborne detection capabilities with process-based, storm-level analysis practices. The MPF database was designed to extend the scale and scope of investigations into the complex connections between precipitation, updrafts, and lightning. The NASA International Space Station Lightning Imaging Sensor (ISS LIS) serves as the source of lightning information for the database. The first iteration of the MPF database leveraged the NASA Global Precipitation Measurement (GPM) mission spaceborne Dual-frequency Precipitation Radar (DPR) to define features, along with contributions of microphysics data and vertical wind retrievals from the GPM Validation Network (VN) of ground-based radar data. The dependency on coincident ISS and GPM satellite overpasses of radars in a dual-Doppler configuration significantly limited the size of the initial database of features, referred to as VNMPFs, but established the database infrastructure and feasibility. We present here a second iteration of the MPF database that omits the GPM DPR and VN, instead incorporating data directly from selected proximal installations of the operational Next Generation Radar (NEXRAD) network that facilitate vertical wind retrievals via dual-Doppler analysis. These features based exclusively on ground-based polarimetric Doppler radar are hereafter referred to as Ground-Radar MPFs (GRMPFs). Removing the restriction of a coincident GPM overpass appreciably increases the size of the GRMPF database while incorporating more detailed information from higher-resolution radar data and retrievals. This expansion allows for unprecedented broad, robust statistical analyses of the electrical, kinematic, and microphysical characteristics of deep convective processes. These results highlight the potential for advancements in lightning meteorology made possible by combining multiple perspectives from global lightning measurements and ground-based radar data.

Lightning↗

Explainable AI (XAI)-driven vibration sensing scheme for surface quality monitoring in a smart surface grinding process

Local Interpretable and Model-agnostic Explanation (LIME), an explainable artificial intelligence (XAI) approach is adapted to identify the globally important time-frequency bands for predicting average surface roughness (Ra) in a smart grinding process. The smart grinding setup consisted of a Supertech CNC precision surface grinding machine, instrumented with a Dytran piezoelectric accelerometer attached to the tailstock quill along the tangential direction (Y-axis). For every grinding pass, vibration signatures were captured, and the ground truth surface roughness values were recorded using a Mahr Marsurf M300C portable surface roughness profilometer. The roughness values ranged from 0.06 to 0.14 microns over the complete set of experiments. Time-frequency domain spectrogram frames were extracted for each of the vibration signals collected during the grinding process. Convolutional Neural Networks (CNNs) were modeled to predict the surface roughness based on these spectrogram frames and their image augmentations. The best CNN model was able to predict the roughness values with an overall R2-score of 0.95, training R2-score of 0.99, and testing R2-score of 0.81 with only 80 sets of vibration signals corresponding to 4 experiments with 20 trials each. Although the data size is not large enough to guarantee such performance metrics in real-world scenarios, one can extract statistically consistent explanations underlying the relationships these complex deep learning models capture. Further, the LIME methodology was implemented on the developed surface roughness CNN model to identify the important time-frequency bands (i.e., the superpixels of a spectrogram) influencing the predictions. Based on the identified important regions on the spectrogram frames, the corresponding frequency characteristics were determined that influence the surface roughness predictions. The important frequency range based on LIME results was approximately 11.7 to 19.1 kHz. The power of XAI was demonstrated by cutting down the sampling rate from 160 kHz to 30, 20, 10, and 5 kHz based on the important frequency range and considering Nyquist criteria. Separate CNN models were developed for these ranges by only extracting time-frequency contents below their corresponding Nyquist cut-offs. A proper data acquisition strategy is proposed by comparing the model performances to argue the selection of a sufficient sampling rate to capture the grinding process successfully and robustly.

42 ENGINEERING↗

Reprint of: Explainable AI (XAI)-driven vibration sensing scheme for surface quality monitoring in a smart surface grinding process

Local Interpretable and Model-agnostic Explanation (LIME), an explainable artificial intelligence (XAI) approach is adapted to identify the globally important time-frequency bands for predicting average surface roughness (Ra) in a smart grinding process. The smart grinding setup consisted of a Supertech CNC precision surface grinding machine, instrumented with a Dytran piezoelectric accelerometer attached to the tailstock quill along the tangential direction (Y-axis). For every grinding pass, vibration signatures were captured, and the ground truth surface roughness values were recorded using a Mahr Marsurf M300C portable surface roughness profilometer. The roughness values ranged from 0.06 to 0.14 microns over the complete set of experiments. Time-frequency domain spectrogram frames were extracted for each of the vibration signals collected during the grinding process. Convolutional Neural Networks (CNNs) were modeled to predict the surface roughness based on these spectrogram frames and their image augmentations. The best CNN model was able to predict the roughness values with an overall R2-score of 0.95, training R2-score of 0.99, and testing R2-score of 0.81 with only 80 sets of vibration signals corresponding to 4 experiments with 20 trials each. Although the data size is not large enough to guarantee such performance metrics in real-world scenarios, one can extract statistically consistent explanations underlying the relationships these complex deep learning models capture. Further, the LIME methodology was implemented on the developed surface roughness CNN model to identify the important time-frequency bands (i.e., the superpixels of a spectrogram) influencing the predictions. Based on the identified important regions on the spectrogram frames, the corresponding frequency characteristics were determined that influence the surface roughness predictions. The important frequency range based on LIME results was approximately 11.7 to 19.1 kHz. The power of XAI was demonstrated by cutting down the sampling rate from 160 kHz to 30, 20, 10, and 5 kHz based on the important frequency range and considering Nyquist criteria. Separate CNN models were developed for these ranges by only extracting time-frequency contents below their corresponding Nyquist cut-offs. A proper data acquisition strategy is proposed by comparing the model performances to argue the selection of a sufficient sampling rate to capture the grinding process successfully and robustly.

47 OTHER INSTRUMENTATION↗

Surface energy fluxes in complex terrain

The emphasis of the 1985 NASA project activity was on field measurements of wind data and heat balance data. Initiatives included a 19 station mountaintop monitoring program, testing and refining the surface flux monitoring systems and packing and shipping equipment to the People's Republic of China in preparation for the 1986 Tibet Experiment. Other work included more extensive analyses of the 1984 Gobi Desert and Rocky Mountain observations plus some preliminary analyses of the 1985 mountaintop network data. Details of our field efforts are summarized and results of our data analyses are presented.

Reiter, E. R.↗

Transients in Pacific/North American Plate Boundary Deformation: Synthesis and Modeling of GPS and Borehole Strain Observations

This is the Final Technical Report on research conducted between 1 June 1997 and 14 September 2001 entitled "Transients in Pacific/North American plate boundary deformation: Synthesis and modeling of GPS and borehole strain observations." As the project title implies, our effort involved a geodetic study of strain transients, i.e., temporal variations in deformation rates, that occur within plate boundary zones and their relationship to earthquakes and plate motions. Important transients occur during and following large earthquakes, and there are also strain transients not apparently associated with earthquakes. A particularly intriguing class of transients, for which there is a modest but growing list of examples, are preseismic anomalies. Such earthquake precursors, if further documented and understood, would have obvious importance for earthquake hazard mitigation. Because the timescales for these diverse transients range over at least 6 orders of magnitude (minutes to years), no single geodetic technique is optimum. We therefore undertook a systematic synthesis of Global Positioning Satellite (GPS) and borehole strainmeter data in three areas in California where there are adequate numbers of both types of instruments (or their equivalent): the San Francisco Bay region (within the Bay Area Regional Deformation network), southern California (within the Southern California Integrated GPS Network), and Parkfield (where a two-color laser system provides a proxy for continuous GPS measurements). An integral component of our study was the elucidation of the physical mechanisms by which such transients occur and propagate. We therefore initiated the development of multiple forward models, using two independent approaches. In the first, we explored the response to specified earthquake slip in viscoelastic models that incorporated failure criteria and the geometry of major faults in California. In the second approach, we examined the dynamical response of a complex rheological medium to the application of a far-field stress imposed by plate motions. The forward models were used both to gain insight into the range of strain transients to be expected under different assumed mechanical conditions and to develop representations for strain fields that allow GPS, borehole, and other strain data to be combined in a self-consistent, yet well-determined, manner. The models also provided a basis for hypothesis testing, by which data from a strain transient well characterized by GPS and borehole observations were utilized to distinguish among competing candidates for the causative physical mechanism and the governing physical characteristics. During the three years of this project, continued to a fourth year through a no-cost extension of the grant, we published 14 papers and presented or co-authored 37 papers at national scientific meetings.

Solomon, Sean C.↗

Channel Measurements for Switching Strategies in Hybrid RF/Optical Communications

Free space optical (FSO) communication links increase data rate, reduce size and power, and increase security. These criteria are particularly important in space communication. Increasing mission complexity and crowding of lower frequency bands is driving the need for optical communications. This makes FSO communication technology extremely attractive, and there is significant ongoing work towards the development of FSO transceivers, ground stations, and relays. Notable projects include the Laser Communications Relay Demonstrator (LCRD), Integrated LCRD Low-Earth Orbit User Modem and Amplifier Terminal (ILLUMA-T), and CubeSat Laser Infrared Crosslink (CLICK) CubeSats. These technologies will eventually operate in unison with existing Radio Frequency (RF) systems, but there is little experimental investigation of such hybrid networks. This paper presents some experimental underpinnings of switching strategies for hybrid RF/FSO systems in various attenuation conditions. A 170 m optical path was constructed in an enclosed test chamber where atmospheric conditions can be tightly controlled. The performance of a 1550 nm infrared FSO link was evaluated in this chamber under varying conditions of turbulence and jitter. The system will eventually be used to investigate switching criteria between the FSO and RF channels. Optimizing the use of RF/FSO communication links will allow data rate, size, power, and security improvements. Therefore, this research will help to mature the network architecture and improve the performance of communication networks to be used for LEO, GEO, Lagrange, Lunar missions and beyond.

Space Networks↗

Validation of macromolecular flexibility in solution by small-angle X-ray scattering (SAXS)

The dynamics of macromolecular conformations are critical to the action of cellular networks. Solution X-ray scattering studies, in combination with macromolecular X-ray crystallography (MX) and nuclear magnetic resonance (NMR), strive to determine complete and accurate states of macromolecules, providing novel insights describing allosteric mechanisms, supramolecular complexes, and dynamic molecular machines. This review addresses theoretical and practical concepts, concerns, and considerations for using these techniques in conjunction with computational methods to productively combine solution-scattering data with high-resolution structures. I discuss the principal means of direct identification of macromolecular flexibility from SAXS data followed by critical concerns about the methods used to calculate theoretical SAXS profiles from high-resolution structures. The SAXS profile is a direct interrogation of the thermodynamic ensemble and techniques such as, for example, minimal ensemble search (MES), enhance interpretation of SAXS experiments by describing the SAXS profiles as population-weighted thermodynamic ensembles. I discuss recent developments in computational techniques used for conformational sampling, and how these techniques provide a basis for assessing the level of the flexibility within a sample. Although these approaches sacrifice atomic detail, the knowledge gained from ensemble analysis is often appropriate for developing hypotheses and guiding biochemical experiments. Examples of the use of SAXS and combined approaches with X-ray crystallography, NMR, and computational methods to characterize dynamic assemblies are presented.

59 BASIC BIOLOGICAL SCIENCES↗

Generalization error guaranteed auto-encoder-based nonlinear model reduction for operator learning

Many physical processes in science and engineering are naturally represented by operators between infinite-dimensional function spaces. The problem of operator learning, in this context, seeks to extract these physical processes from empirical data, which is challenging due to the infinite or high dimensionality of data. An integral component in addressing this challenge is model reduction, which reduces both the data dimensionality and problem size. In this paper, we utilize low-dimensional nonlinear structures in model reduction by investigating Auto-Encoder-based Neural Network (AENet). AENet first learns the latent variables of the input data and then learns the transformation from these latent variables to corresponding output data. Our numerical experiments validate the ability of AENet to accurately learn the solution operator of nonlinear partial differential equations. Furthermore, we establish a mathematical and statistical estimation theory that analyzes the generalization error of AENet. Finally, our theoretical framework shows that the sample complexity of training AENet is intricately tied to the intrinsic dimension of the modeled process, while also demonstrating the robustness of AENet to noise.

Auto-encoder↗

A learning-based data-driven forecast approach for predicting future reservoir performance

Quantification of the predictive uncertainty of subsurface models has long been investigated. The traditional workflow is to calibrate prior models to match observed data, and then use the posterior models to simulate future system performance. Not only are these procedures computationally expensive, but they also have issues in maintaining geological model constraints during the calibration step. Data space inversion (DSI) was introduced recently to predict future system performance without the iterative history matching or model calibration step. In general, DSI approaches seek to establish a statistical relationship between the observed and forecast variables, as well as to quantify the predictive uncertainty of the forecast variables, by using an ensemble of uncalibrated prior models. Existing DSI approaches all require a number of complex transformation and mapping operations, which may deter their widespread use. Here in this study, we introduce a new and simpler DSI approach, the learning-based, data-driven forecast approach (LDFA), by combining dimension reduction and machine learning techniques to quickly provide accurate forecast results and reliably quantify corresponding uncertainty in the results. Our LDFA framework is demonstrated using two supervised learning algorithms, artificial neural network (ANN) and support vector regression (SVR), on two representative examples from reservoir engineering and geological carbon storage. Results suggest that our approach provides accurate forecast results (e.g., future oil production rate or cumulative injected CO 2 ) and reasonable predictive uncertainty intervals. Our framework is generic and may be applied to other surface and subsurface problems.

42 ENGINEERING↗