Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “architecture optimization”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 1,225 records · Page 68
All sourcesNASA NTRSDOE OSTI
NASA NTRS1984

Parallel, iterative solution of sparse linear systems: Models and architectures

A model of a general class of asynchronous, iterative solution methods for linear systems is developed. In the model, the system is solved by creating several cooperating tasks that each compute a portion of the solution vector. A data transfer model predicting both the probability that data must be transferred between two tasks and the amount of data to be transferred is presented. This model is used to derive an execution time model for predicting parallel execution time and an optimal number of tasks given the dimension and sparsity of the coefficient matrix and the costs of computation, synchronization, and communication. The suitability of different parallel architectures for solving randomly sparse linear systems is discussed. Based on the complexity of task scheduling, one parallel architecture, based on a broadcast bus, is presented and analyzed.

Reed, D. A.↗
NASA NTRS1999

Tensegrity and mechanoregulation: from skeleton to cytoskeleton

OBJECTIVE: To elucidate how mechanical stresses that are applied to the whole organism are transmitted to individual cells and transduced into a biochemical response. DESIGN: In this article, we describe fundamental design principles that are used to stabilize the musculoskeletal system at many different size scales and show that these design features are embodied in one particular form of architecture that is known as tensegrity. RESULTS: Tensegrity structures are characterized by use of continuous tension and local compression; architecture, prestress (internal stress prior to application of external force), and triangulation play the most critical roles in terms of determining their mechanical stability. In living organisms, use of a hierarchy of tensegrity networks both optimizes structural efficiency and provides a mechanism to mechanically couple the parts with the whole: mechanical stresses applied at the macroscale result in structural rearrangements at the cell and molecular level. CONCLUSION: Due to use of tensegrity architecture, mechanical stress is concentrated and focused on signal transducing molecules that physically associate with cell surface molecules that anchor cells to extracellular matrix, such as integrins, and with load-bearing elements within the internal cytoskeleton and nucleus. Mechanochemical transduction may then proceed through local stress-dependent changes in molecular mechanics, thermodynamics, and kinetics within the cell. In this manner, the entire cellular response to stress may be orchestrated and tuned by altering the prestress in the cell, just as changing muscular tone can alter mechanical stability and structural coordination throughout the whole musculoskeletal system.

NASA Discipline Cell Biology↗
DOE OSTI2021

Highly accurate and constrained density functional obtained with differentiable programming

Using an end-to-end differentiable implementation of the Kohn-Sham self-consistent field equations, we obtain a highly accurate neural network–based exchange and correlation (XC) functional of the electronic density. The functional is optimized using information on both energy and density while exact constraints are enforced through an appropriate neural network architecture. Here we evaluate our model against different families of XC approximations and show that at the meta-GGA level our functional exhibits unprecedented accuracy for both energy and density predictions. For nonempirical functionals, there is a strong linear correlation between energy and density errors. We use this correlation to define an XC functional quality metric that includes both energy and density errors, leading to an improved way to rank different approximations.

36 MATERIALS SCIENCE↗
DOE OSTI2021

Modeling Data Movement Performance on Heterogeneous Architectures

The cost of data movement on parallel systems varies greatly with machine architecture, job partition, and nearby jobs. Performance models that accurately capture the cost of data movement provide a tool for analysis, allowing for communication bottlenecks to be pinpointed. Modern heterogeneous architectures yield increased variance in data movement as there are a number of viable paths for inter-GPU communication. In this paper, we present performance models for the various paths of inter-node communication on modern heterogeneous architectures, including the trade-off between GPUDirect communication and copying to CPUs. Furthermore, we present a novel optimization for inter-node communication based on these models, utilizing all available CPU cores per node. Finally, we show associated performance improvements for MPI collective operations.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2024

A Performance and Energy Study of GPU-Resident Preconditioners for Conjugate Gradient Solvers: In the Context of Existing and Novel Approaches

Optimizing a particular subprogram out of the set of Basic (sparse) Linear Algebra Subprograms (BLAS) for a given architecture is a common topic of research. In applications, however, these BLAS functions rarely appear in isolation; usually, many of them are used together, in various combinations and with varying inputs. As the need to solve a large, sparse linear system is ubiquitous throughout HPC applications, linear solvers constitute a realistic, sufficiently complex and well-defined representative use case for composite BLAS routines. To this end, based on a representative set of matrices drawn from a diverse set of fields, we present a framework to study, from the performance and energy perspective, the efficacy of GPU- resident parallel Conjugate Gradient (CG) linear solver with different preconditioner options, including Gauss-Seidel, Jacobi, and incomplete Cholesky. We also propose a novel GPU-based preconditioner, in which the triangular solves are approximated by an iterative process. The development of this preconditioner was motivated by solving large graph Laplacian linear systems, for which the existing preconditioners either perform slow on GPU-based platforms or are not applicable. We compare the performance of these preconditioners on different hardware accelerator architectures, i.e., AMD MI250X, MI100, Nvidia A100, V100, and Jetson. Our experiments reveal performance trade-offs and provide information on how to select the best strategy for the given linear system, dictated by its properties, and the platform of interest. We demonstrate the application of our novel preconditioner for solving CG and graph Laplacian systems. Overall, the framework can be utilized as a benchmark to guide informed decisions in choosing a specific preconditioner, i.e., whether it is better to rely on the performance of a triangular solver or on the performance of sparse matrix-vector product. Finally, by considering power consumption to solve the linear systems, we report the energy footprint for the solvers.

Preconditioned Conjugate Gradient, GPUs, iterative↗
DOE OSTI2025

AI-Powered Knowledge Graphs for Neuromorphic and Energy-Efficient Computing

The surge in scientific literature obscures breakthroughs and hinders the discovery of new research paths. We propose an artificial intelligence (AI) powered framework using large language models (LLMs) and knowledge graphs (KGs) to automate parts of scientific discovery, focusing on energy-efficient AI circuits. Our hybrid approach combines LLMs, structured data, and ontology-based reasoning to construct a comprehensive knowledge graph that integrates insights across computational neuroscience, spiking neuron models, learning rules, architectural motifs, and neuromorphic device technologies. This multi-domain representation enables the generation of hypotheses that connect biological function with implementable, energy-efficient hardware architectures. Using KG embeddings and graph neural networks, the framework generates hypotheses for novel circuits, validates them through optimization on exascale HPC systems, and with tools like SuperNeuro and Fugu, the most promising designs will be prototyped in hardware. This open-source system aims to accelerate discoveries and bridging neuroscience with hardware innovation, drive collaboration, and unlock new opportunities in low-power AI computing.

Gautam, Ashish [ORNL]↗
DOE OSTI2025

Structural features of xylan dictate reactivity and functionalization potential for bio-based materials

Plant-based materials have the potential to replace some petroleum-based products, offering compostability and biodegradability as critical advantages. Xylan-rich biomass sources are gaining recognition due to their abundance and underutilization in current industrial applications. Research of potential xylan applications has been complicated by the complex and heterogeneous structure that varies for different xylan feedstocks. Acylation is a broadly used reaction in functionalization of polysaccharides at an industrial scale. However, the efficiency of this reaction varies with the xylan source. To optimize xylan valorization, a systematic understanding of structure–reactivity relationships is essential. This study explores, characterizes, and compares various xylan feedstocks in the acylation process. Xylan feedstocks were analyzed for their chemical composition, degree of polymerization, branching, solubility, and presence of impurities. These features were correlated with xylan glycotypes’ reactivity toward functionalization with succinic anhydride in an optimized DMSO/KOH condition, achieving carboxyl contents of up to 1.46. We used principal component analysis and hierarchical clustering to identify key structural features of xylan that promote its reactivity. Our findings reveal that xylans with higher xylose content and lower degrees of branching exhibit enhanced reactivity, achieving higher carboxyl content and yields. Structural analyses confirmed successful modification, and light scattering analyses showed dramatic changes in the solution properties. Succinylation improves the solubility and film-forming properties of native xylans. This study shows key structure–reactivity relationships in xylan succinylation, establishing that low branching, high xylose content, and reduced lignin impurity enhance chemical functionalization. The results offer a framework for selecting optimal biomass feedstocks and support future efforts in genetic and synthetic biology to design plants with tunable xylan architectures. These findings advance the hemicellulose valorization for applications in coatings and packaging.

Acylation↗
DOE OSTI2016

Digital Architecture Planning Model

As part of the U.S. Department of Energy’s Light Water Reactor Sustainability Program, the Digital Architecture (DA) Project focuses on providing a model that nuclear utilities can refer to when planning deployment of advanced technologies. The digital architecture planning model (DAPM) is the methodology for mapping power plant operational and support activities into a DA that unifies all data sources needed by the utilities to operate their plants. The DA is defined as a collection of information technology capabilities needed to support and integrate a wide spectrum of real-time digital capabilities for performance improvements of nuclear power plants. DA can be thought of as integration of the separate instrumentation and control and information systems already in place in nuclear power plants, which are brought together for the purpose of creating new levels of automation in plant work activities. A major objective in DAPM development was to survey all key areas that needed to be reviewed in order for a utility to make knowledgeable decisions regarding needs and plans to implement a DA at the plant. The development was done in two steps. First, researchers surveyed the nuclear industry in order to learn their near-term plans for adopting new advanced capabilities and implementing a network (i.e., wireless and wire) infrastructure throughout the plant, including the power block. Secondly, a literature review covering regulatory documents, industry standards, and technical research reports and articles was conducted. The objective of the review was to identify key areas to be covered by the DAPM, which included the following: 1. The need for a DA and its benefits to the plant 2. Resources required to implement the DA 3. Challenges that need to be addressed and resolved to implement the DA 4. Roles and responsibilities of the DA implementation plan. The DAPM was developed based on results from the survey and the literature review. Model development, including the survey results and conclusions made about the key areas during the literature review, are described in this report.

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗
DOE OSTI2025

Autonomous Flow Electrochemistry for Accelerated Catalyst Discovery

Our objective is to develop an Autonomous Chemical Experimentation (ACE) platform that accelerates discovery of new catalytic transformations and other energy-relevant chemical reactions and processes. We intentionally designed ACE to be highly modular, both with respect to its rapid deployment to different chemistries and experimental workflows as well as incorporation of a wide range of different AI algorithms. In addition to the development of the core software architecture, initial efforts were made to incorporate Large Language Models to provide human-interpretable reasoning of the optimizer’s actions, and to develop a user-friendly graphical interface for experimental researchers. ACE was demonstrated using a flow electrocatalysis platform containing an inline FTIR spectrometer for real-time analysis and quantification of the reaction outcome. Human-in-the-loop experiments were performed in which a human researcher conducted an experiment using electrode potentials suggested by ACE, then fed the spectral data back to ACE for decision making. After confirming the successful function of the optimizer, efforts were next directed to automation of the hardware and performed full autonomy tests using three reactions: catalytic oxidation of formate, catalytic oxidation of cyclohexanol, and oxidation of hydroquinone. These studies confirm that ACE can close the loop between reaction execution, analysis, and optimization. They also reveal that more improved product detection methods will be essential for ACE to make well-informed decisions for reactions with low conversions.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2026

Exploring the Origins of Anti-Ambipolarity in BBL Polymer: Links to Redox Chemistry, Electronic Structure, and Structural Dynamics

We examine the intrinsic physical-chemical properties of the conjugated ladder-type polymer poly(benzimidazobenzophenanthroline) (BBL) in response to electron transfer. We aim at explaining the origin of the anti-ambipolar behavior behind the observed BBL nonlinear response associated with specific device architectures. To elucidate this point, we use theory and computation based on first principles, including density functional theory optimizations, ab initio molecular dynamics, time-dependent DFT, and Marcus-theory analysis. Our results reveal that this redox response is not simply monotonic but follows an alternating odd/even pattern in which gap narrowing and reopening occur sequentially before near-gapless behavior emerges at high charging. Converging theoretical evidence in this work demonstrates that bell shaped conductivity in BBL originates in its fundamental electronic structure and supramolecular organization.

FOS: Physical sciences↗
NASA NTRS1987

Shape optimization of three-dimensional stamped and solid automotive components

The shape optimization of realistic, 3-D automotive components is discussed. The integration of the major parts of the total process: modeling, mesh generation, finite element and sensitivity analysis, and optimization are stressed. Stamped components and solid components are treated separately. For stamped parts a highly automated capability was developed. The problem description is based upon a parameterized boundary design element concept for the definition of the geometry. Automatic triangulation and adaptive mesh refinement are used to provide an automated analysis capability which requires only boundary data and takes into account sensitivity of the solution accuracy to boundary shape. For solid components a general extension of the 2-D boundary design element concept has not been achieved. In this case, the parameterized surface shape is provided using a generic modeling concept based upon isoparametric mapping patches which also serves as the mesh generator. Emphasis is placed upon the coupling of optimization with a commercially available finite element program. To do this it is necessary to modularize the program architecture and obtain shape design sensitivities using the material derivative approach so that only boundary solution data is needed.

Botkin, M. E.↗
NASA NTRS1989

Systolic array processing of the sequential decoding algorithm

A systolic array processing technique is applied to implementing the stack algorithm form of the sequential decoding algorithm. It is shown that sorting, a key function in the stack algorithm, can be efficiently realized by a special type of systolic arrays known as systolic priority queues. Compared to the stack-bucket algorithm, this approach is shown to have the advantages that the decoding always moves along the optimal path, that it has a fast and constant decoding speed and that its simple and regular hardware architecture is suitable for VLSI implementation. Three types of systolic priority queues are discussed: random access scheme, shift register scheme and ripple register scheme. The property of the entries stored in the systolic priority queue is also investigated. The results are applicable to many other basic sorting type problems.

Chang, C. Y.↗
NASA NTRS1991

Parallel optimization algorithms and their implementation in VLSI design

Two new parallel optimization algorithms based on the simplex method are described. They may be executed by a SIMD parallel processor architecture and be implemented in VLSI design. Several VLSI design implementations are introduced. An application example is reported to demonstrate that the algorithms are effective.

Lee, G.↗
NASA NTRS1995

Artificial neural network application for space station power system fault diagnosis

This study presents a methodology for fault diagnosis using a Two-Stage Artificial Neural Network Clustering Algorithm. Previously, SPICE models of a 5-bus DC power distribution system with assumed constant output power during contingencies from the DDCU were used to evaluate the ANN's fault diagnosis capabilities. This on-going study uses EMTP models of the components (distribution lines, SPDU, TPDU, loads) and power sources (DDCU) of Space Station Alpha's electrical Power Distribution System as a basis for the ANN fault diagnostic tool. The results from the two studies are contrasted. In the event of a major fault, ground controllers need the ability to identify the type of fault, isolate the fault to the orbital replaceable unit level and provide the necessary information for the power management expert system to optimally determine a degraded-mode load schedule. To accomplish these goals, the electrical power distribution system's architecture can be subdivided into three major classes: DC-DC converter to loads, DC Switching Unit (DCSU) to Main bus Switching Unit (MBSU), and Power Sources to DCSU. Each class which has its own electrical characteristics and operations, requires a unique fault analysis philosophy. This study identifies these philosophies as Riddles 1, 2 and 3 respectively. The results of the on-going study addresses Riddle-1. It is concluded in this study that the combination of the EMTP models of the DDCU, distribution cables and electrical loads yields a more accurate model of the behavior and in addition yielded more accurate fault diagnosis using ANN versus the results obtained with the SPICE models.

Momoh, James A.↗
NASA NTRS2006

A Robust Scalable Transportation System Concept

This report documents the 2005 Revolutionary System Concept for Aeronautics (RSCA) study entitled "A Robust, Scalable Transportation System Concept". The objective of the study was to generate, at a high-level of abstraction, characteristics of a new concept for the National Airspace System, or the new NAS, under which transportation goals such as increased throughput, delay reduction, and improved robustness could be realized. Since such an objective can be overwhelmingly complex if pursued at the lowest levels of detail, instead a System-of-Systems (SoS) approach was adopted to model alternative air transportation architectures at a high level. The SoS approach allows the consideration of not only the technical aspects of the NAS", but also incorporates policy, socio-economic, and alternative transportation system considerations into one architecture. While the representations of the individual systems are basic, the higher level approach allows for ways to optimize the SoS at the network level, determining the best topology (i.e. configuration of nodes and links). The final product (concept) is a set of rules of behavior and network structure that not only satisfies national transportation goals, but represents the high impact rules that accomplish those goals by getting the agents to "do the right thing" naturally. The novel combination of Agent Based Modeling and Network Theory provides the core analysis methodology in the System-of-Systems approach. Our method of approach is non-deterministic which means, fundamentally, it asks and answers different questions than deterministic models. The nondeterministic method is necessary primarily due to our marriage of human systems with technological ones in a partially unknown set of future worlds. Our goal is to understand and simulate how the SoS, human and technological components combined, evolve.

Hahn, Andrew↗
NASA NTRS2014

The Use of Software Agents for Autonomous Control of a DC Space Power System

In order to enable manned deep-space missions, the spacecraft must be controlled autonomously using on-board algorithms. A control architecture is proposed to enable this autonomous operation for an spacecraft electric power system and then implemented using a highly distributed network of software agents. These agents collaborate and compete with each other in order to implement each of the control functions. A subset of this control architecture is tested against a steadystate power system simulation and found to be able to solve a constrained optimization problem with competing objectives using only local information.

autonomous control↗
NASA NTRS2016

Trade-Space Analysis Tool for Constellations (TAT-C)

Traditionally, space missions have relied on relatively large and monolithic satellites, but in the past few years, under a changing technological and economic environment, including instrument and spacecraft miniaturization, scalable launchers, secondary launches as well as hosted payloads, there is growing interest in implementing future NASA missions as Distributed Spacecraft Missions (DSM). The objective of our project is to provide a framework that facilitates DSM Pre-Phase A investigations and optimizes DSM designs with respect to a-priori Science goals. In this first version of our Trade-space Analysis Tool for Constellations (TAT-C), we are investigating questions such as: How many spacecraft should be included in the constellation? Which design has the best costrisk value? The main goals of TAT-C are to: Handle multiple spacecraft sharing a mission objective, from SmallSats up through flagships, Explore the variables trade space for pre-defined science, cost and risk goals, and pre-defined metrics Optimize cost and performance across multiple instruments and platforms vs. one at a time.This paper describes the overall architecture of TAT-C including: a User Interface (UI) interacting with multiple users - scientists, missions designers or program managers; an Executive Driver gathering requirements from UI, then formulating Trade-space Search Requests for the Trade-space Search Iterator first with inputs from the Knowledge Base, then, in collaboration with the Orbit Coverage, Reduction Metrics, and Cost Risk modules, generating multiple potential architectures and their associated characteristics. TAT-C leverages the use of the Goddard Mission Analysis Tool (GMAT) to compute coverage and ancillary data, streamlining the computations by modeling orbits in a way that balances accuracy and performance.TAT-C current version includes uniform Walker constellations as well as Ad-Hoc constellations, and its cost model represents an aggregate model consisting of Cost Estimating Relationships (CERs) from widely accepted models. The Knowledge Base supports both analysis and exploration, and the current GUI prototype automatically generates graphics representing metrics such as average revisit time or coverage as a function of cost.

Science Data Processing↗