Search NASA⌕ Search

SEARCH · Search NASA

Results for “temporal learning”

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 469 records · Page 26
All sourcesNASA NTRSDOE OSTI
DOE OSTI2023

Lagrangian large eddy simulations via physics-informed machine learning

High-Reynolds number homogeneous isotropic turbulence (HIT) is fully described within the Navier–Stokes (NS) equations, which are notoriously difficult to solve numerically. Engineers, interested primarily in describing turbulence at a reduced range of resolved scales, have designed heuristics, known as large eddy simulation (LES). LES is described in terms of the temporally evolving Eulerian velocity field defined over a spatial grid with the mean-spacing correspondent to the resolved scale. This classic Eulerian LES depends on assumptions about effects of subgrid scales on the resolved scales. Here, we take an alternative approach and design LES heuristics stated in terms of Lagrangian particles moving with the flow. Our Lagrangian LES, thus L-LES, is described by equations generalizing the weakly compressible smoothed particle hydrodynamics formulation with extended parametric and functional freedom, which is then resolved via Machine Learning training on Lagrangian data from direct numerical simulations of the NS equations. The L-LES model includes physics-informed parameterization and functional form, by combining physics-based parameters and physics-inspired Neural Networks to describe the evolution of turbulence within the resolved range of scales. The subgrid-scale contributions are modeled separately with physical constraints to account for the effects from unresolved scales. We build the resulting model under the differentiable programming framework to facilitate efficient training. We experiment with loss functions of different types, including physics-informed ones accounting for statistics of Lagrangian particles. We show that our L-LES model is capable of reproducing Eulerian and unique Lagrangian turbulence structures and statistics over a range of turbulent Mach numbers.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2023

Reinforcement learning based hybrid bond-order coarse-grained interatomic potentials for exploring mesoscale aggregation in liquid–liquid mixtures

Exploring mesoscopic physical phenomena has always been a challenge for brute-force all-atom molecular dynamics simulations. Although recent advances in computing hardware have improved the accessible length scales, reaching mesoscopic timescales is still a significant bottleneck. Coarse-graining of all-atom models allows robust investigation of mesoscale physics with a reduced spatial and temporal resolution but preserves desired structural features of molecules, unlike continuum-based methods. Here, we present a hybrid bond-order coarse-grained forcefield (HyCG) for modeling mesoscale aggregation phenomena in liquid–liquid mixtures. The intuitive hybrid functional form of the potential offers interpretability to our model, unlike many machine learning based interatomic potentials. We parameterize the potential with the continuous action Monte Carlo Tree Search (cMCTS) algorithm, a reinforcement learning (RL) based global optimizing scheme, using training data from all-atom simulations. The resulting RL-HyCG correctly describes mesoscale critical fluctuations in binary liquid–liquid extraction systems. cMCTS, the RL algorithm, accurately captures the mean behavior of various geometrical properties of the molecule of interest, which were excluded from the training set. The developed potential model along with the RL-based training workflow could be applied to explore a variety of other mesoscale physical phenomena that are typically inaccessible to all-atom molecular dynamics simulations.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2025

Multi-modality deep learning for pulse prediction in homogeneous nonlinear systems via parametric conversion

In this Letter, we introduce FusionNet, a multi-modality deep learning framework designed to predict and analyze output pulses in high-power rare-earth-doped laser systems driving parametric conversion in homogeneous guided nonlinear media. FusionNet integrates temporal, spectral, and physical experimental conditions to model ultrafast nonlinear phenomena, including parametric nonlinear frequency conversion, self-phase modulation, and cross-phase modulation in homogeneous guided systems such as gas-filled hollow-core fibers. These systems bridge physical models with experimental data, advancing our understanding of light-guiding principles and nonlinear interactions while expediting the design and optimization of on-demand high-power, high-brightness systems. Our results demonstrate a 73% reduction in prediction error and an 83% improvement in computational efficiency compared to conventional neural networks. This work establishes a new paradigm for accelerating parametric simulations and optimizing experimental designs in high-power laser systems, with further implications for high-precision spectroscopy, quantum information science, and distributed entangled interconnects.

47 OTHER INSTRUMENTATION↗
DOE OSTI2017

Big Data-Based Approach to Detect, Locate, and Enhance the Stability of an Unplanned Microgrid Islanding

In this paper, a big data-based approach is proposed for the security improvement of an unplanned microgrid islanding (UMI). The proposed approach contains two major steps: the first step is big data analysis of wide-area monitoring to detect a UMI and locate it; the second step is particle swarm optimization (PSO)-based stability enhancement for the UMI. First, an optimal synchrophasor measurement device selection (OSMDS) and matching pursuit decomposition (MPD)-based spatial-temporal analysis approach is proposed to significantly reduce the volume of data while keeping appropriate information from the synchrophasor measurements. Second, a random forest-based ensemble learning approach is trained to detect the UMI. When combined with grid topology, the UMI can be located. Then the stability problem of the UMI is formulated as an optimization problem and the PSO is used to find the optimal operational parameters of the UMI. An eigenvalue-based multiobjective function is proposed, which aims to improve the damping and dynamic characteristics of the UMI. Finally, the simulation results demonstrate the effectiveness and robustness of the proposed approach.

24 POWER TRANSMISSION AND DISTRIBUTION↗
DOE OSTI2018

Advanced Instrumentation for Transient Reactor Testing

Transient testing involves placing fuel or material into the core of specialized materials test reactors that are capable of simulating a range of design basis accidents, including reactivity insertion accidents, that require the reactor produce short bursts of intense highpower neutron flux and gamma radiation. Testing fuel behavior in a prototypic neutron environment under high-power, accident-simulation conditions is a key step in licensing nuclear fuels for use in existing and future nuclear power plants. Transient testing of nuclear fuels is needed to develop and prove the safety basis for advanced reactors and fuels. In addition, modern fuel development and design increasingly relies on modeling and simulation efforts that must be informed and validated using specially designed material performance separate effects studies. These studies will require experimental facilities that are able to support variable scale, highly instrumented tests providing data that have appropriate spatial and temporal resolution. Finally, there are efforts now underway to develop advanced light water reactor (LWR) fuels with enhanced performance and accident tolerance. These advanced reactor designs will also require new fuel types. These new fuels need to be tested in a controlled environment in order to learn how they respond to accident conditions. For these applications, transient reactor testing is needed to help design fuels with improved performance. In order to maximize the value of transient testing, there is a need for in-situ transient realtime imaging technology (e.g., the neutron detection and imaging system like the hodoscope) to see fuel motion during rapid transient excursions with a higher degree of spatial and temporal resolution and accuracy. There also exists a need for new small, compact local sensors and instrumentation that are capable of collecting data during transients (e.g., local displacements, temperatures, thermal conductivity, neutron flux, etc.).

22 GENERAL STUDIES OF NUCLEAR REACTORS↗
DOE OSTI2025

Graph-Based Prediction of Spatio-Temporal Vaccine Hesitancy From Insurance Claims Data

Growing vaccine hesitancy is contributing to the decline in immunization rates for highly contagious, vaccine-preventable childhood diseases. Therefore, there has been a significant interest in understanding how hesitancy is spreading at higher spatio-temporal resolutions, enabling more targeted interventions. Motivated by this, we study the problem of prediction of vaccine hesitancy at the ZIP Code level, referred to as the VaxHesitancy problem. A significant challenge for this problem is the lack of high-resolution data that indicates hesitancy. Here, we develop a hybrid VaxHesSTL framework that combines a Graph Neural Network (GNN) and a Recurrent Neural Network (RNN) to address the VaxHesitancy problem. The GNN uses a ZIP Code-level network to capture spatial signals from neighboring areas, while the RNN models the temporal dynamics present in the data. We train and evaluate VaxHesSTL using a large dataset, namely the All-Payer Claims Databases (APCD), for Virginia, consisting of insurance claims from over five million individuals for six years. We find that an aggregated contact network or graph, developed from a detailed activity-based population network, plays an important role in the performance of VaxHesSTL, compared to graph models based solely on spatial proximity. Experiments demonstrate that VaxHesSTL outperforms a range of state-of-the-art baselines, which rely solely on historical time series data without accounting for spatial relationships. Since hesitancy data at higher spatial resolution is often unavailable or hard to get, we incorporate an active learning approach with our VaxHesSTL framework to optimize the training set without compromising the prediction performance. We find that hesitancy data for only 18% of ZIP Codes selected by active learning allows us to forecast hesitancy for all the ZIP Codes in the Virginia.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2021

A Physics-Constrained Deep Learning Model for Simulating Multiphase Flow in 3D Heterogeneous Porous Media

Physics-based simulators for multiphase flow in porous media emulate nonlinear processes with coupled physics, and usually require extensive computational resources for software development, maintenance and simulation execution. As a result, a huge demand exists for fast modeling of coupled processes in a wide range of subsurface applications including geological sequestration, hydrocarbon recovery and geothermal energy extraction. In this work, an efficient physics-constrained deep learning model is developed for solving multiphase flow in 3-Dimensional (3D) heterogeneous porous media. The model fully leverages the spatial topology predictive capability of convolutional neural networks, specifically U-Net with successive contracting and expansive steps, and is coupled with an efficient continuity-based smoother to predict flow responses that need spatial continuity. Furthermore, the transient regions are penalized to steer the training process such that the model can accurately capture flow in these regions. The model takes inputs including properties of porous media, fluid properties and well controls, and predicts the temporal-spatial evolution of the state variables (pressure and saturation). While maintaining the continuity of fluid flow, the 3D spatial domain is decomposed into 2D images for reducing training cost, and the decomposition results in an increased number of training data samples and better training efficiency. Additionally, a surrogate model is separately constructed as a postprocessor to calculate well flow rate based on the predictions of state variables from the deep learning model. We use the example of CO 2 injection into saline aquifers, and apply the physics-constrained deep learning model that is trained from physics-based simulation data and emulates the physics process. The model performs prediction with a speedup of ~ 1400 times compared to physics-based simulations, and the average temporal errors of predicted pressure and saturation plumes are 0.27% and 0.099% respectively. Furthermore, water production rate is efficiently predicted by a surrogate model for well flow rate, with a mean error less than 5%. Therefore, with its unique scheme to cope with the fidelity in fluid flow in porous media, the physics-constrained deep learning model can become an efficient predictive model for computationally demanding inverse problems or other coupled processes.

58 GEOSCIENCES↗
DOE OSTI2021

A HPC Theory-Guided Machine Learning Cyberinfrastructure for Communicating Hydrometeorological Data Across Scales

High-resolution predictions of hydrometeorological variables are critical for supporting hydropower generation decisions and flood control at hydroelectric power plants. Traditional climate and hydrologic models rely on the numerical simulation of detailed physical processes. Therefore, running these simulations is time-, labor-, and computation-intensive. Improving the spatial and temporal resolution in these modeling outputs could lead to cubic increases in both the simulation time and computational demands, rendering high-resolution hydrometeorological predictions expensive and impractical. Many past studies apply the super resolution (SR) technique to downscale climate models using deep learners. However, deep learners are deemed “black-boxes,” as their derivation processes from low-resolution outputs to high-resolution outputs are often hidden. Their results are difficult for domain scientists to interpret and validate. Thus, there is a need for an exploratory machine learning approach that can partially integrate domain-specific theory and knowledge into the data-driven mapping process between simulation outputs of different spatial scales. The domain-specific theory and knowledge can be incorporated into the data model through an inductive approach in which process-related environmental variables are used and analyzed as key drivers (i.e., environmental surrogates) to reflect the complex physical processes. Many of these variables, such as land use land cover, soil types, topography, digital elevation, air temperature, and various watershed characteristics, can be directly measured through sensors or remote sensing techniques. Additionally, SR applications that can downscale hydrological and hydrodynamics models to efficiently produce high-resolution (1 m) flood depth grids are still rare. Since the flood depth grid can be used to support critical decisions for flood control operation at hydroelectric power plants, it is crucial to enable an SR-based capability for interpolating high-resolution flood inundation maps.

13 HYDRO ENERGY↗
DOE OSTI2025

Multi-Omics Reveals Temporal Scales of Carbon Metabolism in Synechococcus Elongatus PCC 7942 Under Light Disturbance

Central carbon metabolism in model cyanobacteria involves multiple pathways to adapt to energy-light limitations across diel cycles. However, the success in mechanistic modeling for phenotypic prediction of the protein regulators in the metabolic state depends on capturing the vast possibilities emerging from multiple regulatory pathways in complex biological processes. Here, we developed a physics-informed machine learning approach based on energy-landscape concepts to predict regulatory proteins responding to cyclic circadian and unforeseen light perturbations in cyanobacterial metabolic networks. Our approach provides interpretable de novo models for inferring gene expression dynamics from Synechococcus elongatus over diel cycles and using redox proteome analysis to distinguish immediate light-responsive elements from circadian-regulated processes in carbon metabolism pathways. We identified distinct temporal signatures with the analysis of the redox proteome: there was an immediate shift in cysteine redox states accompanied by a limited change in protein abundance under constant illumination and after 2 hours of darkness. This discovery indicates that the generation of reductants coordinates photoinduced electron transport with redox metabolic pathways in two discernable molecular mechanisms: fast redox-based protein modifications occur immediately after the light disturbance, followed by slow transcriptional regulations across networks. This temporal regulation reveals how metabolic networks integrate rapid light responses with programmed circadian rhythms to maintain cellular homeostasis under the light-energy limitations over the diel cycle.

Biomolecular & subcellular processes↗
DOE OSTI2023

Identifying and Explaining Anomalous Activity in Surveillance Video with Compression Algorithms

The primary purpose of this document is to outline the progress made on the LDRD titled “Identifying and Explaining Anomalous Activity in Surveillance Video with Compression Algorithms” in FY22 and FY23. In this LDRD, we explored the usage of compression-based analytics to identify anomalous activity in video. We developed a novel algorithm, Spatio-Temporal N-Gram PPM (STNG PPM) that accounts for spatially and temporally aware anomalies in video. We extracted features using motions vectors from video as well as operating on the raw features. STNG PPM is comparable to many deep learning approaches but does not require specialized hardware (GPUs) to run efficiently. We also examine the evaluation metrics and propose novel measures addressing faults in the current evaluation measures.

97 MATHEMATICS AND COMPUTING↗
NASA NTRS1992

Progress in unstructured-grid methods development for unsteady aerodynamic applications

The development of unstructured-grid methods for the solution of the equations of fluid flow and what was learned over the course of the research are summarized. The focus of the discussion is on the solution of the time-dependent Euler equations including spatial discretizations, temporal discretizations, and boundary conditions. An example calculation with an implicit upwind method using a CFL number of infinity is presented for the Boeing 747 aircraft. The results were obtained in less than one hour CPU time on a Cray-2 computer, thus, demonstrating the speed and robustness of the capability. Additional calculations for the ONERA M6 wing demonstrate the accuracy of the method through the good agreement between calculated results and experimental data for a standard transonic flow case.

Batina, John T.↗
DOE OSTI2018

Technical Advances for Sustainable Remediation: In Situ Monitoring and Climate Resiliency - 18436

Recent advances in sustainable remediation techniques show promise as solutions for contaminant removal in soil and groundwater media. Although passive remediation (or natural attenuation) has gained momentum as an acceptable technique, there are still challenges in regulatory frameworks as well as in public acceptance. Leaving contaminants in subsurface requires increased burden of proof to show that plumes are stable and residual contaminants do not pose a significant health risk. Climate changes and frequent extreme climate events contribute another level of complexity and make it difficult to demonstrate system stability and predict future plume conditions. In this study, we show two coupled approaches to tackle this challenge: (1) in situ groundwater monitoring based on advanced sensor and data analytics technologies (i.e., Big Data), and (2) state-of-art modeling for optimizing monitoring and site closure strategies. In the new monitoring approach, in situ sensors are used to continuously measure the key variables that control contaminant plume mobility and its spatial and temporal distribution (such as pH, redox potential, electrical conductivity, and groundwater level). Coupled with a limited number of groundwater sampling, the data analytics methods - data mining and machine learning - allow us to estimate contaminant concentrations, and also to detect any significant changes associated with the plume mobility. In parallel, the state-of-art numerical flow and reactive transport model - ASCEM software under DOE Office of Environmental Management - provides a physical and mechanistic understanding of the system behavior, and predicts the long-term plume migration for optimizing the monitoring and closure strategies. In particular, we focus on the effects of extreme precipitation or drought on the site closure as well as monitoring strategies. We demonstrate this approach at the Savannah River Site (SRS) F-Area, where groundwater is contaminated by various radionuclides, including uranium, tritium and technetium. (authors)

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗
DOE OSTI2023

Using mechanistic models and machine learning to design single-color multiplexed nascent chain tracking experiments

mRNA translation is the ubiquitous cellular process of reading messenger-RNA strands into functional proteins. Over the past decade, large strides in microscopy techniques have allowed observation of mRNA translation at a single-molecule resolution for self-consistent time-series measurements in live cells. Dubbed Nascent chain tracking (NCT), these methods have explored many temporal dynamics in mRNA translation uncaptured by other experimental methods such as ribosomal profiling, smFISH, pSILAC, BONCAT, or FUNCAT-PLA. However, NCT is currently restricted to the observation of one or two mRNA species at a time due to limits in the number of resolvable fluorescent tags. In this work, we propose a hybrid computational pipeline, where detailed mechanistic simulations produce realistic NCT videos, and machine learning is used to assess potential experimental designs for their ability to resolve multiple mRNA species using a single fluorescent color for all species. Our simulation results show that with careful application this hybrid design strategy could in principle be used to extend the number of mRNA species that could be watched simultaneously within the same cell. We present a simulated example NCT experiment with seven different mRNA species within the same simulated cell and use our ML labeling to identify these spots with 90% accuracy using only two distinct fluorescent tags. We conclude that the proposed extension to the NCT color palette should allow experimentalists to access a plethora of new experimental design possibilities, especially for cell Signaling applications requiring simultaneous study of multiple mRNAs.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2022

Implementing Spatio-Temporal 3D-Convolution Neural Networks and UAV Time Series Imagery to Better Predict Lodging Damage in Sorghum

Unmanned aerial vehicle (UAV)-based remote sensing is gaining momentum in a variety of agricultural and environmental applications. Very-high-resolution remote sensing image sets collected repeatedly throughout a crop growing season are becoming increasingly common. Analytical methods able to learn from both spatial and time dimensions of the data may allow for an improved estimation of crop traits, as well as the effects of genetics and the environment on these traits. Multispectral and geometric time series imagery was collected by UAV on 11 dates, along with ground-truth data, in a field trial of 866 genetically diverse biomass sorghum accessions. We compared the performance of Convolution Neural Network (CNN) architectures that used image data from single dates (two spatial dimensions, 2D) versus multiple dates (two spatial dimensions + temporal dimension, 3D) to estimate lodging detection and severity. Lodging was detected with 3D-CNN analysis of time series imagery with 0.88 accuracy, 0.92 Precision, and 0.83 Recall. This outperformed the best 2D-CNN on a single date with 0.85 accuracy, 0.84 Precision, and 0.76 Recall. The variation in lodging severity was estimated by the best 3D-CNN analysis with 9.4% mean absolute error (MAE), 11.9% root mean square error (RMSE), and goodness-of-fit (R2) of 0.76. This was a significant improvement over the best 2D-CNN analysis with 11.84% MAE, 14.91% RMSE, and 0.63 R2. The success of the improved 3D-CNN analysis approach depended on the inclusion of “before and after” data, i.e., images collected on dates before and after the lodging event. The integration of geometric and spectral features with 3D-CNN architecture was also key to the improved assessment of lodging severity, which is an important and difficult-to-assess phenomenon in bioenergy feedstocks such as biomass sorghum. This demonstrates that spatio-temporal CNN architectures based on UAV time series imagery have significant potential to enhance plant phenotyping capabilities in crop breeding and Precision agriculture applications.

3D-convolution neural networks↗
DOE OSTI2021

Machine Learning and Economic Models to Enable Risk-Informed Condition Based Maintenance of a Nuclear Plant Asset

The primary objective of this research is to address challenges in the implementation of risk-informed, condition-based predictive maintenance (PdM), which reduces operating costs while still maintaining the safety and reliability of commercial nuclear power plants (NPPs). To achieve the objective, risk models are being developed by taking advantage of advancements in data analytics, deep learning, machine learning (ML), and artificial intelligence (AI). The notable outcomes presented in the report include ? Development of a ML models using heterogeneous plant process and vibration data collected at different spatial and temporal resolutions from the Salem?s CWS to diagnose a circulating water pump (CWP) failure based on salient fault signatures. The developed diagnostic models are extendable to other faults associated with CWPs and CWP motors given associated fault signatures. ? Development of a natural language processing (NLP) technique to automatically classify the WO data into different categories. The developed NLP technique was validated on independent WO data. This automates the tedious and time-consuming activity of mining and classifying WOs by subject matter experts. ? Estimation of mean time between downtime (i.e., time duration between time instances when 1 or more CWPs are not available) and developed an approach to establish reliability of CWS components using unstructured WO data along with CWS plant process data. ? Formulation of economic model based on Markov chain models. The parameters of associated with the transition rate between different states of Markov chain models were estimated using WO data. The economic model formulation and discussion captures both time-independent and time-dependent parameter variation, leading to risk-informed decision-making.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗
DOE OSTI2025

Mapping wall-to-wall fractional cover of Arctic tundra plant functional types in Alaska using 20-m spatial resolution satellite imagery and harmonized plot observations

Estimates of fractional cover (fCover) across given land surfaces are used to assess, and often model, vegetation composition and diversity, which are crucial for understanding the health and functioning of terrestrial ecosystems. Remote sensing provides a useful means for scaling local, plot-measured fCover estimates to regional scales. Leveraging a recently synthesized and harmonized plot database, this study generated wall-to-wall maps of fCover for six Alaskan-Arctic plant functional types (PFT), including non-vascular plants, forbs, graminoids, and deciduous and evergreen shrubs, using 20-m satellite data (Sentinel-1, Sentinel-2, ArcticDEM) using a machine learning regression approach, specifically the random forest (RF) algorithm, which is well-suited for handling nonlinear relationships and high-dimensional satellite datasets. This study additionally addressed the spatio-temporal inconsistencies e.g., sampling scale, plot size, and collection year in plot measured fCover by adopting a multivariate outlier detection approach—Cook’s distance—to identify high-quality plots for model training and validation. Our approach achieves high accuracy (R 2 = 0.59–0.93, root mean squared errors = 0.02–0.10 for all PFTs) between plot-observed and satellite-derived fCover when using high-quality plot samples. The mapped fCover characterizes the spatial patterns of different PFTs across the tundra biome at a 20-m resolution, providing key information needed for improved representation of Arctic tundra vegetation in terrestrial biosphere models to better understand climate-vegetation feedback across the Arctic tundra.

Arctic tundra↗
NASA NTRS2007

What can be Learned from X-Ray Spectroscopy Concerning Hot Gas in the Local Bubble and Charge Exchange Processes

Solar wind charge exchange produces diffuse X-ray emission with a variable surface brightness comparable to that of the cosmic background. While the temporal variation of the charge exchange emission allows some separation of the components, there remains a great deal of uncertainty as to the zero level of both. Because the production mechanisms of the two components are considerably different, their spectra would provide critical diagnostics to the understanding of both. However, current X-ray observatories are very limited in both spectral resolution and sensitivity in the critical soft X-ray (less than 1.0 keV) energy range. Non-dispersive high-resolution spectrometers, such as the calorimeter proposed for the Spectrum Roentgen Gamma mission, will be extremely useful in distinguishing the cascade emission of charge exchange from the spectra of thermal bremsstrahlung cosmic plasmas.

Snowden, Steven L.↗
DOE OSTI2019

Optimal Proactive Vehicle Relocation for On-Demand Mobility Service with Deep Convolution-LSTM Network

This study develops an innovative model for inferring optimal proactive relocation strategy for empty vehicles of on-demand mobility service (OMS) directly from past observations of passenger demand. The output of our model is a row stochastic relocation matrix which gives the probability that an empty vehicle should be relocated from location i to location j at any given time. The matrix itself is high-dimensional and therefore challenging to obtain accurate estimation with traditional machine learning approach. Our model proposes the low-rank approximation of the relocation matrix, and introduces the deep convolution-LSTM network which extracts low-level spatial and temporal features from past demand observations for best restoration of the actual low rank approximation vector. The model is trained by deriving the optimal matching strategies from one year of real-world trip data. The results show that our model is able to predict the low-rank approximation of the relocation matrix with 4% of mean absolute error, may reduce the relocation cost by 20%-45% as compared to the real-world observations and is 50% optimal as compared to the global optimal matching with perfect information.

Lei, Zengxiang↗