Search NASA⌕ Search

SEARCH · Search NASA

Results for “load allocation”

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 181 records · Page 10
All sourcesNASA NTRSDOE OSTI
DOE OSTI2021

Enabling Scalable and Extensible Memory-mapped Datastores in Userspace

Exascale workloads are expected to incorporate data-intensive processing in close coordination with traditional physics simulations. These emerging scientific, data-analytics and machine learning applications need to access a wide variety of datastores in flat files and structured databases. Programmer productivity is greatly enhanced by mapping datastores into the application process's virtual memory space to provide a unified “in-memory” interface. Currently, memory mapping is provided by system software primarily designed for generality and reliability. However, scalability at high concurrency is a formidable challenge on exascale systems. Also, there is a need for extensibility to support new datastores potentially requiring HPC data transfer services. In this article, we present UMap , a scalable and extensible userspace service for memory-mapping datastores. Furthermore, through decoupled queue management, concurrency aware adaptation, and dynamic load balancing, UMap enables application performance to scale even at high concurrency. We evaluate UMap in data-intensive applications, including sorting, graph traversal, database operations, and metagenomic analytics. Our results show that UMap as a userspace service outperforms an optimized kernel-based service across a wide range of intra-node concurrency by 1.22-1.9 × . We performed two case studies to demonstrate UMap 's extensibility. First, a new datastore residing in remote memory is incorporated into UMap as an application-specific plugin. Second, we present a persistent memory allocator Metall built atop UMap for unified storage/memory.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2026

matsim-agents v1.0

matsim-agents is a multi-agent AI framework for atomistic materials simulation and discovery. It orchestrates large language models (LLMs), machine-learned interatomic potentials (MLIPs), and DFT codes into a single agentic loop running on laptops and DOE leadership-class supercomputers. MULTI-AGENT ORCHESTRATION A LangGraph state machine with three nodes: a Planner that converts a natural-language research objective into structured tasks; an Executor that dispatches atomistic tools and loops until the queue is empty; and an Analyst that summarizes results into a human-readable report. State is checkpointed after every step and human-in-the-loop gates can be inserted at any edge. HYPOTHESIS-DRIVEN DISCOVERY CHAT An interactive REPL (matsim-agents chat) that couples LLM dialogue with atomistic simulation. Chemical formulas are automatically detected in conversation turns and trigger a full crystal-phase exploration: structure generation → relaxation → stability scoring → result injection back into the conversation, creating a closed hypothesis-refinement loop. CRYSTAL PHASE ENUMERATION Given a composition, the phase explorer enumerates prototypes by stoichiometry: elemental (fcc/bcc/hcp/sc/diamond), binary 1:1 (rocksalt/CsCl/zincblende/ wurtzite/fluorite/rutile), ternary 1:1:3 (cubic perovskite), ternary 1:2:4 (perovskite + spinel), quaternary 1:1:2:6 (Fm-3m double perovskite). 2-D prototypes (graphene, h-BN, MoS2 2H/1T) and multilayer stacking are also supported via --include-2d and --num-layers. SUPERCELL GENERATION AND SITE DECORATION Auto-tiling to a minimum atom count (--min-atoms), explicit NxNxN tiling (--supercell), symmetry-distinct site decorations (--n-orderings), and isotropic lattice-scale sweeps (--lattice-scales) for volume bracketing. MLFF RELAXATION AND STABILITY SCORING HydraGNN (multi-headed GNN) drives structure relaxation via ASE with FIRE, BFGS, or BFGSLineSearch. Stability output: delta-E/atom ranking across phases and a max-residual-force dynamical-stability proxy. Other MLIPs (MACE, NequIP, Orb) can be plugged in through the same interface. DFT BACKENDS Quantum ESPRESSO pw.x and VASP 6.6 are first-class labellers. Both have validated GPU builds and SLURM/PBS launchers for three DOE platforms: Frontier (AMD MI250X, ROCm), Aurora (Intel PVC, oneAPI), Perlmutter (NVIDIA A100, CUDA). QE produces ~100 binaries (pw.x, ph.x, epw.x, ...). VASP supports scf, relax, vc-relax, and vc-relax-shape run types. ACTIVE-LEARNING LOOP matsim-agents al run CONFIG.yaml drives an iterative HydraGNN-DFT loop: MD generates candidates → ensemble/MC-dropout uncertainty selects the most informative → DFT labels them in parallel inside one allocation → dataset grows → HydraGNN retrains → repeat. DFT backend is a single YAML toggle (dft.backend: vasp | qe). LLM-generated seed structures are supported (no curated POSCAR library needed). Config uses ${VAR}, ${VAR:-default}, ${VAR:?msg} shell-style substitution for cross-user/cross-site portability. LLM BACKENDS Ollama (local, default), vLLM (HPC multi-GPU serving), OpenAI, Anthropic, HuggingFace Transformers+Accelerate. Selected at runtime via flag or env var with no code changes. HPC PORTABILITY Same Python entry points run on Frontier (ROCm 7.2), Aurora (oneAPI), and Perlmutter (CUDA 12). DFT and ML stacks are never co-loaded in the same shell; they couple through the scheduler and filesystem. Advanced multi-node launchers (serve, discovery-chat, single-relaxation, active-learning, QE warm-start) are provided for all three platforms. CODABENCH COMPETITION BUNDLE A self-contained benchmark: 159 atomistic test structures across 11 material classes, 5 tasks (formation energy, forces, ML relaxation, AI-DFT relaxation, phase stability ranking), public/private leaderboard split (30/70), and four ready-to-run baselines: MACE-MP-0, HydraGNN, UMA, AllScAIP.

Lupo Pasini, Massimiliano [Oak Ridge National Labo↗
NASA NTRS1974

Space processing applications payload equipment study. Volume 2B: Payload interface analysis (power/thermal/electromagnetic compatibility)

As a part of the task of performing preliminary engineering analysis of modular payload subelement/host vehicle interfaces, a subsystem interface analysis was performed to establish the integrity of the modular approach to the equipment design and integration. Salient areas that were selected for analysis were power and power conditioning, heat rejection and electromagnetic capability (EMC). The equipment and load profiles for twelve representative experiments were identified. Two of the twelve experiments were chosen as being representative of the group and have been described in greater detail to illustrate the evaluations used in the analysis. The shuttle orbiter will provide electrical power from its three fuel cells in support of the orbiter and the Spacelab operations. One of the three shuttle orbiter fuel cells will be dedicated to the Spacelab electrical power requirements during normal shuttle operation. This power supplies the Spacelab subsystems and the excess will be available to the payload. The current Spacelab sybsystem requirements result in a payload allocation of 4.0 to 4.8 kW average (24 hour/day) and 9.0 kW peak for 15 minutes.

Hammel, R. L.↗
DOE OSTI2021

Resilience-Oriented DG Siting and Sizing Considering Stochastic Scenario Reduction

In this paper, a fuel-based distributed generator (DG) allocation strategy is proposed to enhance the distribution system resilience against extreme weather. The long-term planning problem is formulated as a two-stage stochastic mixed-integer programming (SMIP). The first stage is to make decisions of DG siting and sizing under the given budget constraint. In the second stage, a post-extreme-event-restoration (PEER) is employed to minimize the operating cost in an uncertain fault scenario. In particular, this study proposes a method to select the most representative scenarios for the SMIP. First, a Monte Carlo Simulation (MCS) is introduced to generate sufficient scenarios considering random fault locations and load profiles. Then, the number of scenarios is reduced by the K-means clustering algorithm. The advantage of scenario reduction is to make a trade-off between accuracy and computational efficiency. Finally, the SMIP is solved by the progressive hedging algorithm. Here, the case studies of the IEEE 33-bus and 123-bus test systems demonstrate the effectiveness of the proposed algorithm in reducing the expected energy not served (EENS), which is a critical criterion of resilience.

42 ENGINEERING↗
NASA NTRS2007

DSN Resource Scheduling

TIGRAS is client-side software, which provides tracking-station equipment planning, allocation, and scheduling services to the DSMS (Deep Space Mission System). TIGRAS provides functions for schedulers to coordinate the DSN (Deep Space Network) antenna usage time and to resolve the resource usage conflicts among tracking passes, antenna calibrations, maintenance, and system testing activities. TIGRAS provides a fully integrated multi-pane graphical user interface for all scheduling operations. This is a great improvement over the legacy VAX VMS command line user interface. TIGRAS has the capability to handle all DSN resource scheduling aspects from long-range to real time. TIGRAS assists NASA mission operations for DSN tracking of station equipment resource request processes from long-range load forecasts (ten years or longer), to midrange, short-range, and real-time (less than one week) emergency tracking plan changes. TIGRAS can be operated by NASA mission operations worldwide to make schedule requests for the DSN station equipment.

Wang, Yeou-Fang↗
NASA NTRS2018

Evaluation of Multiple Flow Constrained Area Capacity Setting Methods for Collaborative Trajectory Options Program

The purpose of this study was to compare flow constrained area (FCA) capacity setting methods for Collaborative Trajectory Options Program (CTOP) as they pertain to the Integrated Demand Management (IDM) concept. IDM uses flow balancing to manage air traffic across multiple FCAs with a common downstream constraint, as well as constraints at the respective FCA locations. FCA capacity rates can be set manually, but generating capacities for multiple, interdependent FCAs could potentially over-burden a user. A new enhancement to CTOP called the FCA Balance Algorithm (FBA) was developed at NASA Ames Research Center to improve the process of allocating capacity across multiple flow constrained segments in the airspace. The FBA evaluates the predicted demand and capacity across multiple FCAs and dynamically generates capacity settings for the FCAs that best meet capacity limits for all identified constraints. In a human-in-the-loop simulation study, both manual and automated capacity setting methods were evaluated in terms of their overall feasibility using measures of system performance, human performance, and qualitative feedback. Subject matter experts were asked to use three different methods to allocate capacity to three FCAs, either (1) by manually setting capacity for every 60-minute time window, (2) by manually setting capacity for every 15-minute time window, or (3) by using the FBA capability to automatically generate capacity settings. Results showed no significant differences in terms of overall system performance, indicated by similar ground delay and airport throughput numbers between methods. However, differences in individual strategies afforded by the manual methods allowed some participants to achieve system-wide delay that was much lower than the average. The FBA was the fastest method of capacity setting, and it received the lowest subjective rating scores on physical task load, mental task load, task difficulty and task complexity out of the three methods. Finally, participants explained through qualitative feedback that there were many benefits to using the FBA, such as ease of use, accuracy, and low risk of human input error. Participants did not experience the same limitations with the FBA that they did with the manual methods, such as reduced accuracy in the 60-minute manual condition, or high complexity in the 15-minute/manual condition. These results suggest that the FBA automation enhancement to CTOP maintains system performance while improving human performance. Therefore, the FBA could be introduced as a way to mitigate operator workload while planning a CTOP.

NextGen↗
NASA NTRS2018

Evaluation of Multiple Flow Constrained Area Capacity Setting Methods for Collaborative Trajectory Options Program

The purpose of this study was to compare flow constrained area (FCA) capacity setting methods for Collaborative Trajectory Options Program (CTOP) as they pertain to the Integrated Demand Management (IDM) concept. IDM uses flow balancing to manage air traffic across multiple FCAs with a common downstream constraint, as well as constraints at the respective FCA locations. FCA capacity rates can be set manually, but generating capacities for multiple, interdependent FCAs could potentially over-burden a user. A new enhancement to CTOP called the FCA Balance Algorithm (FBA) was developed at NASA Ames Research Center to improve the process of allocating capacity across multiple flow constrained segments in the airspace. The FBA evaluates the predicted demand and capacity across multiple FCAs and dynamically generates capacity settings for the FCAs that best meet capacity limits for all identified constraints. In a human-in-the-loop simulation study, both manual and automated capacity setting methods were evaluated in terms of their overall feasibility using measures of system performance, human performance, and qualitative feedback. Subject matter experts were asked to use three different methods to allocate capacity to three FCAs, either (1) by manually setting capacity for every 60-minute time window, (2) by manually setting capacity for every 15-minute time window, or (3) by using the FBA capability to automatically generate capacity settings. Results showed no significant differences in terms of overall system performance, indicated by similar ground delay and airport throughput numbers between methods. However, differences in individual strategies afforded by the manual methods allowed some participants to achieve system-wide delay that was much lower than the average. The FBA was the fastest method of capacity setting, and it received the lowest subjective rating scores on physical task load, mental task load, task difficulty and task complexity out of the three methods. Finally, participants explained through qualitative feedback that there were many benefits to using the FBA, such as ease of use, accuracy, and low risk of human input error. Participants did not experience the same limitations with the FBA that they did with the manual methods, such as reduced accuracy in the 60-minute manual condition, or high complexity in the 15-minute/manual condition. These results suggest that the FBA automation enhancement to CTOP maintains system performance while improving human performance. Therefore, the FBA could be introduced as a way to mitigate operator workload while planning a CTOP.

NextGen↗
DOE OSTI2022

Enabling DER Participation in Frequency Regulation Markets

Distributed energy resources (DERs) are playing an increasing role in ancillary services for the bulk grid, particularly in frequency regulation. In this article, we propose a framework for collections of DERs, combined to form microgrids and controlled by aggregators, to participate in frequency regulation markets. Our approach covers both the identification of bids for the market clearing stage and the mechanisms for the real-time allocation of the regulation signal. The proposed framework is hierarchical, consisting of a top layer and a bottom layer. The top layer consists of the aggregators communicating in a distributed fashion to optimally disaggregate the regulation signal requested by the system operator. The bottom layer consists of the DERs inside each microgrid whose power levels are adjusted so that the tie line power matches the output of the corresponding aggregator in the top layer. The coordination at the top layer requires the knowledge of cost functions, ramp rates, and capacity bounds of the aggregators. We develop meaningful abstractions for these quantities respecting the power flow constraints and taking into account the load uncertainties and propose a provably correct distributed algorithm for optimal disaggregation of regulation signals among the microgrids.

24 POWER TRANSMISSION AND DISTRIBUTION↗
DOE OSTI2022

6.26 Low Cost Basalt Fiber for Automotive Applications

Vehicle lightweighting is an essential component to the automotive industry to improve fuel economy of internal combustion engine (ICE) vehicles to meet ever improving emission standards and to improve the range of electric vehicles (EV). Composite materials offer high specific modulus and specific strength, which makes them appealing for these light weighting efforts. Sheet molding compounds (SMC) are particularly interesting from an automotive perspective because of the relatively low cost and high volume of producing SMC parts. Traditionally, composite materials for automotive application are glass fiber reinforced because of the attractive price - performance ratio, but basalt fibers are a cost and recycling competitive reinforcement alternative in this market. The aim of this project was to examine the feasibility of utilizing basalt fiber for automotive applications. More specifically, an effort was made to examine different fiber sizings on basalt fiber combined with vinyl-ester (VE) resin, and their performance as part of an SMC process. In addition to offering vehicle lightweighting with fiber reinforced polymer composites, basalt fiber is a fully recyclable material and thus supports the IACMI technical goal of: Demonstrate that the technology is capable at a sufficient scale for >80% recyclability or reuse of fiber reinforced polymer composites in five years into useful components with projected cost and quality at commercial scale competitive with virgin materials on a pathway to 95% recyclability or reuse starting in ten years. Three different fiber sizings were applied to a continuous roving of basalt fiber and compared to a traditional Electrical/Chemical Resistance (E-CR) glass fiber that is typically used in these types of applications. Fiber tows were examined for Loss on Ignition percentage (LOI%), Tex, and tow strength. Some sizings clearly outperformed others, and the ability to process these fibers on a pilot scale SMC line was demonstrated. A test plan for the manufacturing and mechanical testing of SMC panels was developed. This work continues outside the time frame allocated for this project. When this work is completed, it will be added to this report and posted as Appendix C. Glass fiber reinforced SMC materials have already proven feasible as a light weighting method for traditionally steel parts like the Volkswagen (VW) Atlas lift-gate (Figure 1); this project team is seeking the feasibility of basalt fiber as a drop-in replacement for glass fiber reinforced SMC. Sizing development for basalt fibers has proven that the mechanical properties are better than E-glass and closer to S-glass, which makes it an interesting material for SMC applications. Better mechanical properties translate to less material needed to achieve load case requirement for target applications. The business case has already been demonstrated for 100,000 parts per year of glass fiber reinforced SMC Atlas lift-gates compared to traditional steel manufacturing processes. Reduced overhead and assembly costs are offset by glass fiber SMC higher cost per kg beyond 100,000 parts per year, which is still a relatively low volume for the automotive industry. For basalt fiber reinforced SMC to become feasible for automotive applications, the price-performance ratio has to be precisely determined. Based on the mechanical performance it is possible to establish a range of applications and technical solutions in which the potential of basalt SMC can be utilized, while the price of the material can be used to compile the business case for such applications. Based on these business cases and the sustainability indicators, glass fiber reinforcement (or other) materials can be directly substituted. Volkswagen’s commitment to reducing carbon emissions cannot be understated. Basalt fiber shows promise of reducing the carbon footprint in SMC materials, especially if sizing optimizations can be made with thermoplastic based SMC. To fully realize the value of basalt fiber reinforced materials, a lifecycle cost analysis should be performed on basalt’s production and recycling, and then compared against E-glass. From this assessment, a true judgement can be made on the commercialization potential of this material. Figure 1. Example of Fiber Reinforced Polymer Composite Liftgate As a conclusion, we can state that Mafic basalt fiber is not a direct replacement for E-glass or E-CR glass based on price, but should be considered a technical solution when E-glass does not provide adequate performance in a composite design and S-glass, aramid and carbon fibers are too costly. Mafic basalt fiber can be placed on the high-performance fiber spectrum next to S-glass for performance but at one third the price. It should be considered for more technically challenging structural designs wherein the performance can demonstrate 20-25% performance enhancement over E-glass to elicit more strength or a weight reduction. Both Michelman and Mafic produce thermoplastic sizings which, in combination with Nylon and polypropylene resin and fibers, can further advance high speed composite implementations while maintaining an eco-friendly manufacturing process.

36 MATERIALS SCIENCE↗
NASA NTRS2019

DVB-S2 Demonstration Testing for Enhancing Data Rates for CubeSat/SmallSat Missions

The number of National Aeronautics and Space Administration (NASA) CubeSat/SmallSat missions is expected to grow rapidly in the next decade. High data rate is in increasing demanded for science missions, especially for mother/daughter CubeSat constellations. As the number of spacecraft on a ground network grows, loading could be reduced by limiting contact time per day, which is enabled by higher data rates. There is also a need to communicate direct to earth from space from longer distances than low earth orbit (LEO) with CubeSats. These challenges motivate the need for bandwidth and power efficient modulation and coding techniques. Today, DVB-S2 is an industry communications standard for larger satellites. DVB-S2 uses power and bandwidth efficient modulation and coding techniques to deliver performance approaching theoretical limits of RF channels. NASA Near Earth Network (NEN) is conducting demonstration testing at Wallops Flight Facility (WFF) in spring 2019 for CubeSat/SmallSat missions for enhancing data rate performance in NASAs allocated S-band 5 MHz channel. The ultimate goal is to upgrade NEN with DVB-S2 for increasing science data return, and enabling of greater numbers of CubeSats.This paper describes NEN DVB-S2 demonstration testing objectives and performance measurement results. DVB-S2 data rate performance in the NEN S-band 5 MHz channel is presented. Simulation analysis for the expected maximum data rate performance with the DVB-S2 signal family including multiple modulations and codes are presented. Link margin analysis for a typical CubeSat/SmallSat with a 1W/2W power amplifier/patch antenna using DVB-S2 for NEN S-band is discussed. The demonstration testing configurations at NEN Wallops station is described. Results of the demonstration testing are compared with other evolving radios for SmallSats and CubeSats in term of data rate and performance. There are a number of evolving S-band and X-band radios that are compatible with NEN. Some are integrated with commercial CubeSat/SmallSat busses. Some are flying for the first time with NEN in early 2019. The USRP B200mini transceiver S-band radio is one such radio. Results of streamlined compatibility testing with NEN in 2018, performance with the Technology Educational Satellite (TechEdSat-8) in flight in early 2019, potential to add DVB-S2 to the radio, and plans for a CubeSat using the radio for communication from 10 million kilometers from earth are discussed. Another evolving radio is the Syrlinks X-band radio. Results of compatibility testing with NEN in 2018, performance with the Sustained Ocean Color Observations using Nanosatellites (SOCON) in flight in early 2019, and potential for a future SOCON constellation are discussed.

Wong, Yen↗
NASA NTRS1995

Experimental evaluation of dynamic data allocation strategies in a distributed database with changing workloads

Traditionally, allocation of data in distributed database management systems has been determined by off-line analysis and optimization. This technique works well for static database access patterns, but is often inadequate for frequently changing workloads. In this paper we address how to dynamically reallocate data for partionable distributed databases with changing access patterns. Rather than complicated and expensive optimization algorithms, a simple heuristic is presented and shown, via an implementation study, to improve system throughput by 30 percent in a local area network based system. Based on artificial wide area network delays, we show that dynamic reallocation can improve system throughput by a factor of two and a half for wide area networks. We also show that individual site load must be taken into consideration when reallocating data, and provide a simple policy that incorporates load in the reallocation decision.

Brunstrom, Anna↗
NASA NTRS2001

Expression of heat shock protein 72 in atrophied rat skeletal muscles

Changes in the expression of heat shock protein 72 (HSP72) in response to atrophic-inducing perturbations of muscle involving chronic mechanical unloading and denervation were determined. Adult male Wistar rats were assigned randomly to a sedentary cage control (CON), hind limb unloading (HU, via tail suspension), HU plus tenotomy (HU + TEN), HU plus denervation (HU + DEN), or HU + TEN + DEN group. Tenotomy and DEN involved cutting the Achilles tendon and removing a segment of the sciatic nerve, respectively. After 5 days, HSP72 levels in the soleus of the HU + DEN and HU + TEN + DEN groups were 42 (P < 0.05) and 53% (P < 0.01) less than CON, respectively. Soleus weight decreased in both groups. Heat shock protein 72 levels in the plantaris of the HU + TEN, HU + DEN, and HU + TEN + DEN groups were 31, 25, and 30% lower than CON, respectively (P < 0.05). Plantaris weight decreased in the HU + DEN and HU + TEN + DEN, but not in the HU + TEN group. Hind limb unloading alone had little effect on the HSP72 level in either muscle. Reduced levels of HSP72 were associated with a decreased soleus (r=0.62, P < 0.01) and plantaris (r=0.78, P < 0.001) weight. These results indicate that the levels of HSP72 in both a slow and a fast rat plantarflexor are responsive to a chronic decrease in the levels of loading and/or activation and suggest that the neuromuscular activity level and the presence of innervation of a muscle are important factors that induce HSP72 expression.

NASA Discipline Musculoskeletal↗
DOE OSTI2025

Adaptive Computing for Scale-Up Problems

Adaptive Computing is an application-agnostic outer loop framework to strategically deploy simulations and experiments to guide decision making for scale-up analysis. Resources are allocated over successive batches, which makes the allocation adaptive to some objective such as optimization or model training. The framework enables the characterization and management of uncertainties associated with predictive models of complex systems when scale-up questions lead to significant model extrapolation. A key advancement of this framework is its integration of multi-fidelity surrogate modeling, uncertainty management, and automated orchestration of various computing and experimentation resources into a single integrated software package. This enables efficient multi-fidelity modeling across multiple computing resources by incorporating real-world constraints such as relative queue times and throughput on individual machines into the multi-fidelity sampling decision. We discuss applications of this framework to problems in the renewable energy space, including biofuels production, material synthesis, perovskite crystal growth, and building electrical loads.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗
DOE OSTI2020

Network traffic control for multi-homed end-hosts via SDN

Software-defined networking (SDN) is an emerging technology of efficiently controlling and managing computer networks, such as in data centres, wide-area networks, as well as in ubiquitous communication. In this study, the authors explore the idea of embedding the SDN components, represented by SDN controller and virtual switch, in end-hosts to improve network performance. In particular, the authors consider load balancing across multiple network interfaces on end-hosts with different link capacity scenarios. The authors have explored and implemented different SDN-based load-balancing approaches based on OpenFlow software switches, and have demonstrated the feasibility and the potential of this approach. The proposed system has been evaluated with MultiPath transmission control protocol (MPTCP). Furthermore, the proposed results demonstrated the potential of applying the SDN concepts on multi-homed devices resulting in an increase in achieved throughput of 55% compared to the legacy single network approach and 10% compared to the MPTCP.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2021

Hardware-Based Emulator with Deep Learning Model for Building Energy Control and Prediction Based on Occupancy Sensors’ Data

Heating, ventilation, and air conditioning (HVAC) is the largest source of residential energy consumption. Occupancy sensors’ data can be used for HVAC control since it indicates the number of people in the building. HVAC and sensors form a typical cyber-physical system (CPS). In this paper, we aim to build a hardware-based emulation platform to study the occupancy data’s features, which can be further extracted by using machine learning models. In particular, we propose two hardware-based emulators to investigate the use of wired/wireless communication interfaces for occupancy sensor-based building CPS control, and the use of deep learning to predict the building energy consumption with the sensor data. We hypothesize is that the building energy consumption may be predicted by using the occupancy data collected by the sensors, and question what type of prediction model should be used to accurately predict the energy load. Another hypothesis is that an in-lab hardware/software platform could be built to emulate the occupancy sensing process. The machine learning algorithms can then be used to analyze the energy load based on the sensing data. To test the emulator, the occupancy data from the sensors is used to predict energy consumption. The synchronization scheme between sensors and the HVAC server will be discussed. We have built two hardware/software emulation platforms to investigate the sensor/HVAC integration strategies, and used an enhanced deep learning model—which has sequence-to-sequence long short-term memory (Seq2Seq LSTM)—with an attention model to predict the building energy consumption with the preservation of the intrinsic patterns. Because the long-range temporal dependencies are captured, the Seq2Seq models may provide a higher accuracy by using LSTM architectures with encoder and decoder. Meanwhile, LSTMs can capture the temporal and spatial patterns of time series data. The attention model can highlight the most relevant input information in the energy prediction by allocating the attention weights. The communication overhead between the sensors and the HVAC control server can also be alleviated via the attention mechanism, which can automatically ignore the irrelevant information and amplify the relevant information during CNN training. Our experiments and performance analysis show that, compared with the traditional LSTM neural network, the performance of the proposed method has a 30% higher prediction accuracy.

Ye, Zhijing↗
DOE OSTI2024

Efficient Parallelization of Irregular Applications on GPU Architectures

With the enlarging computation capacity of general Graphics Processing Units (GPUs), leveraging GPUs to accelerate parallel applications has become a critical topic in academia and industry. However, a wide range of irregular applications with the computation-/memory-intensive nature cannot easily achieve high GPU utilization. The challenges mainly involve the following aspects: first, data dependence leads to coarse-grained kernel and inefficient parallelism; second, heavy GPU memory usage may cause frequent memory evictions and extra overhead of I/O; third, specific computation patterns produce memory redundancies; last, workload balance and data reusability conjunctly benefit the overall performance, but there may exist a dynamic trade-off between them. Targeting these challenges, this dissertation proposes multiple optimizations to accelerate two real-world applications: many-body correlation functions to simulate nuclear physics in a large-scale scientific system; the other is the eALS-based matrix factorization recommendation system. To accelerate the calculations of many-body correlation functions, this dissertation presents three frameworks in GPU memory management and multi-GPU scheduling. Firstly, an optimized systematic GPU memory management framework, MemHC, utilizes a series of new memory reduction designs in GPU memory allocation, CPU/GPU communications, and GPU memory oversubscription. Secondly, an enhanced multi-GPU scheduling framework, MICCO, particularly by taking both data dimension (e.g., data reuse and data eviction) and computation dimension into account. MICCO designs a heuristic scheduling algorithm and a machine learning-based regression model to generate the optimal settings of a proposed new concept to manage the trade-off. Thirdly, a locality-aware multi-GPU scheduling framework. This scheduler leverages pipeline batch generation with a looking-ahead strategy by building local dependency graphs for memory transfer reduction and better data reuse, achieving up to 79.92% memory cost reduction and 1.67x speedup. To parallelize the eALS-based recommendation system, this dissertation proposes an efficient CPU/GPU heterogeneous recommendation system, HEALS. HEALS employs newly designed architecture-adaptive data formats to achieve load balance and good data locality on CPU and GPU. To mitigate the data dependence, HEALS presents a CPU/GPU collaboration model for both task parallelism and data parallelism with multiple kernel computation optimizations. In summary, this dissertation efficiently accelerates two typical irregular applications on GPUs by building four frameworks, including CPU/GPU collaboration, GPU memory management, and multi-GPU scheduling.

Wang, Qihan↗
NASA NTRS2022

PAAV Concept Document

The Pathfinding for Airspace with Autonomous Vehicles (PAAV) Concept Document, version 1.0, lays out the key challenges and potential solutions for the use of uncrewed aircraft (UA) technology for future regional air cargo operations. The challenges and solutions described in this document were informed by communications with the UA industry community (e.g., RTCA, the Federal Aviation Administration, and regional air cargo business operators), as well as the PAAV team’s research activities during the last two years including four tabletop exercises, a human-in-the-loop simulation study, a numerical simulation study, a functional allocation study, and flight data analysis (Appendix A). This document first describes the expected operational context of PAAV (Section 2), such as the flight mission, baseline UAS components, nominal operations, m:N operations (i.e., "m" remote pilots per "N" aircraft), and off-nominal operations. This context sets the scope for the PAAV concept development work. PAAV concept development assumes that UA operations will be increasingly autonomous. Thus, near- and far-term assumptions are defined (Section 3). PAAV identified seven key challenges for UA operations (Section 4): - Flight route planning - Separation and flow management - Traffic pattern integration - Contingency management - Taxi, takeoff, and landing - m:N operations - Communications operations The following 13 potential solutions to these challenges are then described (Section 5): - Scalable communications architecture - Data link - Designated UAS corridors - Crew planning for m:N operations - Flight route optimization - Traffic load-level control - Trajectory solutions with data link - Automated hazard avoidance for m:N operations - Traffic pattern integration (TPI) tool - Standard lost command and control (C2) link (LC2L) procedures - Automated hazard avoidance under LC2L - Auto-taxi, auto-takeoff, and auto-land - Ground control station (GCS) user interface for m:N operations The document attempts to link each of these solutions to one or more of the challenge areas. Novel solutions involving numerous automation technologies are needed to mitigate traffic and airspace management challenges, especially for realizing m:N operations and ensuring safety under LC2L conditions. The purpose of this document is to help understand alternatives and tradeoffs among potential solutions and provide a foundation for a cohesive PAAV concept that will be described and refined in subsequent concept versions.

Unmanned aircraft, uncrewed aircraft, regional air↗
DOE OSTI2022

Load‐Balancing Intense Physics Calculations to Embed Regionalized High‐Resolution Cloud Resolving Models in the E3SM and CESM Climate Models

Abstract We design a new strategy to load‐balance high‐intensity sub‐grid atmospheric physics calculations restricted to a small fraction of a global climate simulation's domain. We show why the current parallel load balancing infrastructure of Community Earth System Model (CESM) and Energy Exascale Earth Model (E3SM) cannot efficiently handle this scenario at large core counts. As an example, we study an unusual configuration of the E3SM Multiscale Modeling Framework (MMF) that embeds a binary mixture of two separate cloud‐resolving model grid structures that is attractive for low cloud feedback studies. Less than a third of the planet uses high‐resolution (MMF‐HR; sub‐km horizontal grid spacing) relative to standard low‐resolution (MMF‐LR) cloud superparameterization elsewhere. To enable MMF runs with Multi‐Domain cloud resolving models (CRMs), our load balancing theory predicts the most efficient computational scale as a function of the high‐intensity work's relative overhead and its fractional coverage. The scheme successfully maximizes model throughput and minimizes model cost relative to precursor infrastructure, effectively by devoting the vast majority of the processor pool to operate on the few high‐intensity (and rate‐limiting) high‐resolution (HR) grid columns. Two examples prove the concept, showing that minor artifacts can be introduced near the HR/low‐resolution CRM grid transition boundary on idealized aquaplanets, but are minimal in operationally relevant real‐geography settings. As intended, within the high (low) resolution area, our Multi‐Domain CRM simulations exhibit cloud fraction and shortwave reflection convergent to standard baseline tests that use globally homogenous MMF‐LR and MMF‐HR. We suggest this approach can open up a range of creative multi‐resolution climate experiments without requiring unduly large allocations of computational resources.

54 ENVIRONMENTAL SCIENCES↗