Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “network performance evaluation”

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 217 records · Page 12
All sourcesNASA NTRSDOE OSTI
DOE OSTI2025

Investigating the Vulnerabilities of the Direct Transfer Trip Scheme for Network Protector Units in the Secondary Networks of Electric Power Distribution Grids

Network protector units (NPUs) are crucial parts of the protection of secondary networks to effectively isolate faults occurring on the primary feeders. When a fault occurs on the primary feeder, there is a path of the fault current going through the service transformers that causes a negative flow of current on the NPU connected to the faulted feeder. Conventionally, NPUs rely on the direction of current with respect to the voltage to detect faults and make a correct trip decision. However, the conventional NPU logic does not allow the reverse power flow caused by distributed energy resources installed on secondary networks. The communication-assisted direct transfer trip logic for NPUs can be used to address this challenge. However, the communication-assisted scheme is exposed to some vulnerabilities arising from the disruption or corruption of the communicated data that can endanger the reliable operation of NPUs. This paper evaluates the impact of the malfunction of the communication system on the operation of communication-assisted NPU logic. To this end, the impact of packet modification and denial-of-service cyberattacks on the communication-assisted scheme are evaluated. The evaluation was performed using a hardware-in-the-loop (HIL) co-simulation testbed that includes both real-time power system and communication network digital simulators. This paper evaluates the impact of the cyberattacks for different fault scenarios and provides a list of recommendations to improve the reliability of communication-assisted NPU protection.

direct transfer trip↗
NASA NTRS1970

Solar electric propulsion system evaluation

Solar electric propulsion system performance consisting of thrustors with thrust vector aligning actuators, switching network and flight type power conditioner

Costogue, E. N.↗
DOE OSTI2022

Design of an Additively Manufactured Recuperator With 800 °C Inlet Temperature for sCO2 Power Cycle Application

Abstract Recuperators with design temperatures at and above 800 °C can further increase the thermal efficiency of supercritical CO2 power cycles by enabling higher turbine exhaust temperatures. Mar-M247 is a well-suited nickel-based superalloy for high temperature service due to its high creep strength that prevents excessive material thickness being required for pressure containment. Additive manufacturing using a high-speed laser-directed energy deposition (L-DED) process presents a promising solution, with build trials demonstrating the ability to produce nonconventional flow channels for enhanced heat transfer. A design process is presented that includes aerothermal and mechanical evaluation to maximize performance within the constraints of the manufacturing process. A 2-D heat transfer network and pressure drop code allows prediction of flow distribution and its effect on overall thermal performance. Established literature correlations, along with CFD simulation, inform the prediction of heat transfer coefficients and friction factors for the flowpaths and enhancement features in the heat exchanger core. Mechanical evaluation using finite element analysis (FEA) modeling with the intent of the ASME Boiler & Pressure Vessel Code (BPVC) Section VIII, Div. 2 assesses the operational safety of the design. The detailed design features annular finned passages that take advantage of helical flow paths to distribute the flow from separated headers to shared heat transfer surfaces. Performance predictions for the recuperator at a 50 kW scale provide insights into the feasibility of the additively manufactured (AM) process to produce recuperators on a commercial scale that extend existing operating envelopes.

Engineering↗
DOE OSTI2024

Phase Diagrams of Alloys and Their Hydrides via On-Lattice Graph Neural Networks and Limited Training Data

Efficient prediction of sampling-intensive thermodynamic properties is needed to evaluate material performance and permit high-throughput materials modeling for a diverse array of technology applications. To alleviate the prohibitive computational expense of high-throughput configurational sampling with density functional theory (DFT), surrogate modeling strategies like cluster expansion are many orders of magnitude more efficient but can be difficult to construct in systems with high compositional complexity. We therefore employ minimal-complexity graph neural network models that accurately predict and can even extrapolate to out-of-train distribution formation energies of DFT-relaxed structures from an ideal (unrelaxed) crystallographic representation. This enables the large-scale sampling necessary for various thermodynamic property predictions that may otherwise be intractable and can be achieved with small training data sets. Two exemplars, optimizing the thermodynamic stability of low-density high-entropy alloys and modulating the plateau pressure of hydrogen in metal alloys, demonstrate the power of this approach, which can be extended to a variety of materials discovery and modeling problems.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2024

Exploring the Landscape of Distributed Graph Clustering on Leadership Supercomputers

The rapid growth of large-scale datasets in fields like biology and social networks has driven the need for advanced graph analytics techniques. Community detection, a fundamental task in graph analytics, identifies closely connected groups of nodes within a network, providing valuable insights across various disciplines. This study focuses on two classic community detection methods, the Louvain algorithm and Markov Clustering (MCL), and evaluates the performance of two prominent distributed community detection algorithms: HiPDPL-GPU, our prior implementation, and HipMCL. We conduct experiments on GPU-accelerated heterogeneous HPC systems, Summit and Frontier, to assess their performance under varying conditions. Our objective is to identify the strengths and weaknesses of these algorithms in terms of scalability, and quality of solutions. We evaluate these algorithms on a diverse set of 70+ networks spanning 13 domains, with sizes ranging up to 4.2 billion edges. Our results demonstrate that HiPDPL-GPU consistently outperforms HipMCL, especially for large-scale networks. HiPDPL-GPU achieves significantly faster runtimes (47x to 1439x), higher modularity scores, and improved scalability. These findings highlight HiPDPL-GPU as a promising solution for efficient and effective large-scale graph analytics in diverse application domains, and provide insights into the feasibility of using MCL-based approaches for certain application domains.

Community detection, graph algorithms↗
NASA NTRS1997

Astronomical Software Directory Service

With the support of NASA's Astrophysics Data Program (NRA 92-OSSA-15), we have developed the Astronomical Software Directory Service (ASDS): a distributed, searchable, WWW-based database of software packages and their related documentation. ASDS provides integrated access to 56 astronomical software packages, with more than 16,000 URLs indexed for full-text searching. Users are performing about 400 searches per month. A new aspect of our service is the inclusion of telescope and instrumentation manuals, which prompted us to change the name to the Astronomical Software and Documentation Service. ASDS was originally conceived to serve two purposes: to provide a useful Internet service in an area of expertise of the investigators (astronomical software), and as a research project to investigate various architectures for searching through a set of documents distributed across the Internet. Two of the co-investigators were then installing and maintaining astronomical software as their primary job responsibility. We felt that a service which incorporated our experience in this area would be more useful than a straightforward listing of software packages. The original concept was for a service based on the client/server model, which would function as a directory/referral service rather than as an archive. For performing the searches, we began our investigation with a decision to evaluate the Isite software from the Center for Networked Information Discovery and Retrieval (CNIDR). This software was intended as a replacement for Wide-Area Information Service (WAIS), a client/server technology for performing full-text searches through a set of documents. Isite had some additional features that we considered attractive, and we enjoyed the cooperation of the Isite developers, who were happy to have ASDS as a demonstration project. We ended up staying with the software throughout the project, making modifications to take advantage of new features as they came along, as well as influencing the software development. The Web interface to the search engine is provided by a gateway program written in C++ by a consultant to the project (A. Warnock).

Hanisch, Robert J.↗
DOE OSTI2021

Constrained Block Nonlinear Neural Dynamical Models

Neural network modules conditioned by known priors can be effectively trained and combined to represent systems with nonlinear dynamics. This work explores a novel formulation for data-efficient learning of deep control-oriented nonlinear dynamical models by embedding local model structure and constraints. The proposed method consists of neural network blocks that represent input, state, and output dynamics with constraints placed on the network weights and system variables. For handling partially observable dynamical systems, we utilize a state observer neural network to estimate the states of the system's latent dynamics. We evaluate the performance of the proposed architecture and training methods on system identification tasks for three nonlinear systems: a continuous stirred tank reactor, a two tank interacting system, and an aerodynamics body. Models optimized with a few thousand system state observations accurately represent system dynamics in open loop simulation over thousands of time steps from a single set of initial conditions. Experimental results demonstrate an order of magnitude reduction in open-loop simulation mean squared error for our constrained, block-structured neural models when compared to traditional unstructured and unconstrained neural network models.

Skomski, Elliott↗
DOE OSTI2024

An Investigation of LES Wall Modeling for Rayleigh–Bénard Convection via Interpretable and Physics-Aware Feedforward Neural Networks with DNS

Abstract The traditional approach of using the Monin–Obukhov similarity theory (MOST) to model near-surface processes in large-eddy simulations (LESs) can lead to significant errors in natural convection. In this study, we propose an alternative approach based on feedforward neural networks (FNNs) trained on output from direct numerical simulation (DNS). To evaluate the performance, we conduct both a priori and a posteriori tests. In the a priori (offline) tests, we compare the statistics of the surface shear stress and heat flux, computed from filtered DNS input variables, to the stress and flux obtained from the filtered DNS. Additionally, we investigate the importance of various input features using the Shapley additive explanations value and the conditional average of the filter grid cells. In the a posteriori (online) tests, we implement the trained models in the System for Atmospheric Modeling (SAM) LES and compare the LES-generated surface shear stress and heat flux with those in the DNS. Our findings reveal that vertical velocity, a traditionally overlooked flow quantity, is one of the most important input features for determining the wall fluxes. Increasing the number of input features improves the a priori test results but does not always improve the model performance in the a posteriori tests because of the differences in input variables between the LES and DNS. Last, we show that physics-aware FNN models trained with logarithmic and scaled parameters can well extrapolate to more intense convection scenarios than in the training dataset, whereas those trained with primitive flow quantities cannot. Significance Statement The traditional near-surface turbulence model, based on a shear-dominated boundary layer flow, does not represent near-surface turbulence in natural convection. Using a feedforward neural network (FNN), we can construct a more accurate model that better represents the near-surface turbulence in various flows and reveals previously overlooked controlling factors and process interactions. Our study shows that the FNN-generated models outperform the traditional model and highlight the importance of the near-surface vertical velocity. Furthermore, the physics-aware FNN models exhibit the potential to extrapolate to convective flows of various intensities beyond the range of the training dataset, suggesting their broader applicability for more accurate modeling of near-surface turbulence.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2025

Techno-economic implications and cost of forecasting errors in solar PV power production using optimized deep learning models

Accurate solar Photovoltaic (PV) power forecasting is important for enhancing both the performance and economic feasibility of PV systems. This study evaluates several deep learning models, including Dense Neural Networks (DNN), Long Short-Term Memory (LSTM), Convolutional Neural Networks (CNN), and a hybrid LSTMCNN model, for predicting PV power production one day in advance. Prior to optimization, the models exhibited relatively high errors, with the best model (DNN) achieving a Root Mean Square Error (RMSE) of 31.13 kW and a coefficient of determination (R 2 ) of 62.15 %. After employing Bayesian optimization, the LSTM-CNN model demonstrated the best performance, with the RMSE reduced to 9.79 kW and R 2 improved to 97.62 %, showcasing significant enhancement in predictive accuracy. Here, the economic evaluation considered three cases: rewards for underestimation (0.08 USD/kWh), no rewards, and penalties for both over-and underestimation (120 % of the utility tariff). In the rewards scenario, the LSTM-CNN model reduced the Levelized Cost of Electricity (LCOE) by 4 %, while in the penalty scenario, a backup diesel generator would have increased the LCOE by 49 %. Additionally, the LSTM-CNN model minimized financial losses, achieving the lowest penalties and maximizing net cash flow compared to other models, demonstrating its overall technical and economic superiority.

Deep learning↗
NASA NTRS1991

RL-34 ring laser gyro laboratory evaluation for the Deep Space Network antenna application

The overall results of this laboratory evaluation are quite encouraging. The gyro data is in good agreement with the system's overall pointing performance, which is quite close to the technical objectives for the Deep Space Network (DSN) application. The system can be calibrated to the levels required for millidegree levels of pointing performance, and initialization performance is within the required 0.001 degree objective. The blind target acquisition performance is within a factor of two of the 0.0001 degree objective, limited only by a combination of the slow rate (0.5 deg/sec) and the existing production quantization logic (0.38 arc-sec/pulse). Logic circuitry exists to better this performance such that it will better the objective by 50 percent. Representative data with this circuitry has been provided for illustration. Target tracking performance is about twice the one millidegree objective, with several factors contributing. The first factor is the bias stability of the gyros, which is exceptional, but will limit performance to the 0.001 and 0.002 degree range for long tracking periods. The second contributing factor is the accelerometer contributions when the system is elevated. These degrade performance into the 0.003 to 0.004 degree range, which could be improved upon with some additional changes. Finally, we have provided a set of recommendations to improve performance closer to the technical objectives. These recommendations include gyro, electronics, and system configurational changes that form the basis for additional work to achieve the desired performance. In conclusion, we believe that the RL-34 ring laser gyro-based advanced navigation system demonstrated performance consistent with expectations and technical objectives, and it has the potential for even further enhancement for the DSN application.

Source record↗
NASA NTRS2013

A Validated Set of MIDAS V5 Task Network Model Scenarios to Evaluate Nextgen Closely Spaced Parallel Operations Concepts

The Closely Spaced Parallel Operations (CSPO) scenario is a complex, human performance model scenario that tested alternate operator roles and responsibilities to a series of off-nominal operations on approach and landing (see Gore, Hooey, Mahlstedt, Foyle, 2013). The model links together the procedures, equipment, crewstation, and external environment to produce predictions of operator performance in response to Next Generation system designs, like those expected in the National Airspaces NextGen concepts. The task analysis that is contained in the present report comes from the task analysis window in the MIDAS software. These tasks link definitions and states for equipment components, environmental features as well as operational contexts. The current task analysis culminated in 3300 tasks that included over 1000 Subject Matter Expert (SME)-vetted, re-usable procedural sets for three critical phases of flight; the Descent, Approach, and Land procedural sets (see Gore et al., 2011 for a description of the development of the tasks included in the model; Gore, Hooey, Mahlstedt, Foyle, 2013 for a description of the model, and its results; Hooey, Gore, Mahlstedt, Foyle, 2013 for a description of the guidelines that were generated from the models results; Gore, Hooey, Foyle, 2012 for a description of the models implementation and its settings). The rollout, after landing checks, taxi to gate and arrive at gate illustrated in Figure 1 were not used in the approach and divert scenarios exercised. The other networks in Figure 1 set up appropriate context settings for the flight deck.The current report presents the models task decomposition from the tophighest level and decomposes it to finer-grained levels. The first task that is completed by the model is to set all of the initial settings for the scenario runs included in the model (network 75 in Figure 1). This initialization process also resets the CAD graphic files contained with MIDAS, as well as the embedded operator models that comprise MIDAS. Following the initial settings, the model progresses to begin the first tasks required of the two flight deck operators, the Captain (CA) and the First Officer (FO). The task sets will initialize operator specific settings prior to loading all of the alerts, probes, and other events that occur in the scenario. As a note, the CA and FO were terms used in developing this model but the CA can also be thought of as the Pilot Flying (PF), while the FO can be considered the Pilot-Not-Flying (PNF)or Pilot Monitoring (PM). As such, the document refers to the operators as PFCA and PNFFO respectively.

task network model↗
DOE OSTI2024

Identifying Potential Sub-Synchronous Oscillations Using Impedance Scan Approach: Preprint

This paper presents an impedance scan study on the real-world power system in Australia that has observed 17 to 20 Hz intermittent subsynchronous oscillations. Through the impedance scan of each of the IBRs, both individually and collectively, potential resonance modes are identified. The impedance scans were carried out using electromagnetic transient PSCAD models of the network. The network comprises of site-specific, black-boxed models of IBRs supplied by the generators. The impedance scan approach was divided in three major steps: (1) the first step is to identify IBRs where impedance analysis needs to be performed based on the magnitude of oscillations observed at their points of interconnection (POIs); (2) the second step focuses on performing impedance scans at selected IBRs in single-machine infinite-bus (SMIB) configuration to identify internal resonance modes of an IBR and to evaluate if any of these modes become unstable under certain grid conditions; (3) the third step performs impedance scans at a few selected IBRs while connected to the wider network model to obtain the impedance response of both the IBR and the grid - this step evaluates control interactions among IBRs. The impedance scan study found that a few IBRs have an underdamped resonance mode at around 17 Hz, which becomes unstable under a certain operating condition. Another interesting finding was that certain IBRs increase the effective grid impedance seen by another IBR in proximity for a particular operating condition. Under this operating condition, the resulting subsynchronous oscillation mode is more pronounced due to a combination of control interactions among IBRs through the transmission network under study and a resonance mode inside the IBRs.

control interactions↗
DOE OSTI2021

Clinical knowledge extraction via sparse embedding regression (KESER) with multi-center large scale electronic health record data

The increasing availability of electronic health record (EHR) systems has created enormous potential for translational research. However, it is difficult to know all the relevant codes related to a phenotype due to the large number of codes available. Traditional data mining approaches often require the use of patient-level data, which hinders the ability to share data across institutions. In this project, we demonstrate that multi-center large-scale code embeddings can be used to efficiently identify relevant features related to a disease of interest. We constructed large-scale code embeddings for a wide range of codified concepts from EHRs from two large medical centers. We developed knowledge extraction via sparse embedding regression (KESER) for feature selection and integrative network analysis. We evaluated the quality of the code embeddings and assessed the performance of KESER in feature selection for eight diseases. Besides, we developed an integrated clinical knowledge map combining embedding data from both institutions. The features selected by KESER were comprehensive compared to lists of codified data generated by domain experts. Features identified via KESER resulted in comparable performance to those built upon features selected manually or with patient-level data. The knowledge map created using an integrative analysis identified disease-disease and disease-drug pairs more accurately compared to those identified using single institution data. Analysis of code embeddings via KESER can effectively reveal clinical knowledge and infer relatedness among codified concepts. KESER bypasses the need for patient-level data in individual analyses providing a significant advance in enabling multi-center studies using EHR data.

96 KNOWLEDGE MANAGEMENT AND PRESERVATION↗
DOE OSTI2026

wa-hls4ml: A Benchmark and Surrogate Models for hls4ml Resource and Latency Estimation

As machine learning (ML) is increasingly implemented in hardware to address real-time challenges in scientific applications, the development of advanced toolchains has significantly reduced the time required to iterate on various designs. These advancements have solved major obstacles, but also exposed new challenges. For example, processes that were not previously considered bottlenecks, such as hardware synthesis, are becoming limiting factors in the rapid iteration of designs. To mitigate these emerging constraints, multiple efforts have been undertaken to develop an ML-based surrogate model that estimates resource usage of ML accelerator architectures. We introduce wa-hls4ml, a benchmark for ML accelerator resource and latency estimation, and its corresponding initial dataset of over 680,000 fully connected and convolutional neural networks, all synthesized using hls4ml and targeting Xilinx FPGAs. The benchmark evaluates the performance of resource and latency predictors against several common ML model architectures, primarily originating from scientific domains, as exemplar models, and the average performance across a subset of the dataset. Additionally, we introduce GNN- and transformer-based surrogate models that predict latency and resources for ML accelerators. We present the architecture and performance of the models and find that the models generally predict latency and resources for the 75% percentile within several percent of the synthesized resources on the synthetic test dataset.

Hawks, Benjamin [Fermilab] (ORCID:0000000157000288↗
DOE OSTI2025

wa-hls4ml: A Benchmark and Surrogate Models for hls4ml Resource and Latency Estimation

As machine learning (ML) is increasingly implemented in hardware to address real-time challenges in scientific applications, the development of advanced toolchains has significantly reduced the time required to iterate on various designs. These advancements have solved major obstacles, but also exposed new challenges. For example, processes that were not previously considered bottlenecks, such as hardware synthesis, are becoming limiting factors in the rapid iteration of designs. To mitigate these emerging constraints, multipleefforts have been undertaken to develop an ML-based surrogate model that estimates resource usage of synthesized ML accelerator architectures. We introduce wa-hls4ml, a benchmark for ML accelerator resource and latency estimation, and its corresponding initial dataset of over 680 000 fully connected and convolutional neural networks, all synthesized using hls4ml and targeting Xilinx FPGAs. The benchmark evaluates the performance of resource and latency predictors against several common ML model architectures, primarily originating from scientific domains, as exemplar models, and the average performance across a subset of the dataset. Additionally, we introduce GNN- and transformer-based surrogate models that predict latency and resources for ML accelerators. We present the architecture and performance of the models and find that the models generally predict latency and resources for the 75% percentile within several percent of the synthesized resources on the synthetic test dataset.

Hawks, Benjamin G. [Fermilab]↗
DOE OSTI2024

Design, Detection, and Countermeasure of Frequency Spectrum Attack and Its Impact on Long Short-Term Memory Load Forecasting and Microgrid Energy Management

This paper introduces a frequency-domain false data injection attack called Frequency Spectrum Attack (FSA) and explores its effects on load forecasting and the energy management system (EMS) in a microgrid. The FSA analyzes time-series signals in the frequency domain to identify patterns in their frequency spectrum. It learns the distribution of dominant frequencies in a dataset of healthy signals. Subsequently, it manipulates the amplitudes of dominant frequencies within this healthy distribution, ensuring a stealthy attack against statistical analysis of the signal spectrum. We evaluated the performance of FSA on LSTM, a state-of-the-art network for load forecasting. The results show that FSA can triple the Mean Absolute Error (MAE) of predictions compared to the normal case and increase it by 70% compared to noise injection attacks. Furthermore, FSA indirectly enhances battery utilization in the EMS by 45%. We then proposed a detection method that combines statistical analysis and machine-learning-based classification techniques with features. The model effectively distinguishes FSA from healthy and noisy signals, achieving an accuracy of 98.7% and an F1-score of 98.1% on a load dataset, covering healthy, FSA, and noisy load data. Finally, a countermeasure was introduced based on the statistical analysis of the frequency spectrum of healthy signals to mitigate the impact of FSA. This countermeasure successfully reduces the MAE of the attacked model from 0.135 to 0.053, validating its effectiveness in mitigating FSA.

Nazeri, Amirhossein↗
NASA NTRS1971

Automatic transmission line monitor

Monitor improves complex network reliability in computer data links and command transmission lines. System evaluates circuit performance against preselected criteria, identifies and stores data indicating out-of-tolerance conditions, conducts closed loop testing, and provides for operation under command of digital computer that determines restoration priorities.

Parsons, W. E.↗