Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Computational network”

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 1,567 records · Page 87
All sourcesNASA NTRSDOE OSTI
DOE OSTI2024

Complexity-calibrated benchmarks for machine learning reveal when prediction algorithms succeed and mislead

Abstract Recurrent neural networks are used to forecast time series in finance, climate, language, and from many other domains. Reservoir computers are a particularly easily trainable form of recurrent neural network. Recently, a “next-generation” reservoir computer was introduced in which the memory trace involves only a finite number of previous symbols. We explore the inherent limitations of finite-past memory traces in this intriguing proposal. A lower bound from Fano’s inequality shows that, on highly non-Markovian processes generated by large probabilistic state machines, next-generation reservoir computers with reasonably long memory traces have an error probability that is at least $$\sim 60\%$$ ∼ 60 % higher than the minimal attainable error probability in predicting the next observation. More generally, it appears that popular recurrent neural networks fall far short of optimally predicting such complex processes. These results highlight the need for a new generation of optimized recurrent neural network architectures. Alongside this finding, we present concentration-of-measure results for randomly-generated but complex processes. One conclusion is that large probabilistic state machines—specifically, large $$\epsilon$$ ϵ -machines—are key to generating challenging and structurally-unbiased stimuli for ground-truthing recurrent neural network architectures.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2020

Improved pore network models to simulate single-phase flow in porous media by coupling with lattice Boltzmann method

In this paper, different pore network models to simulate single-phase flow in porous media are built and their accuracy is evaluated. In addition to the conventional pore network model (CPNM) which consists of regular pore bodies and throat bonds, three improved pore network models (IPNMs) are developed allowing to better describing the real pore and throat geometry. The first improved pore network model (IPNM1) replaces the regular throat bond with a throat bond showing the real throat cross section. The second improvement (IPNM2) uses a series of sub-throat bonds with varying cross sections to better describe the real throat geometry, which is firstly proposed in this paper. The third model (IPNM3) extracts the real pore-throat-pore geometry without simplification. The conductance of fluid flow through these more realistic throat bonds is calculated by the lattice Boltzmann method (LBM). The accuracy and computational efficiency of the different pore network models are evaluated taking the LBM simulation over the whole porous medium as reference solution. The global permeability and detailed pressure distributions in the pores for the different pore network models are validated. The results show that the accuracy of the pore network model increases from CPNM to IPNM3, but at the expense of increasing computational cost. This study suggests that IPNM3 can replace a whole-domain LBM simulation with similar accuracy but much lower computational cost. As a first-order approximation the newly proposed IPNM2 is suggested as good compromise between accuracy and computational cost.

42 ENGINEERING↗
NASA NTRS1981

Management and development of local area network upgrade prototype

Given the situation of having management and development users accessing a central computing facility and given the fact that these same users have the need for local computation and storage, the utilization of a commercially available networking system such as CP/NET from Digital Research provides the building blocks for communicating intelligent microsystems to file and print services. The major problems to be overcome in the implementation of such a network are the dearth of intelligent communication front-ends for the microcomputers and the lack of a rich set of management and software development tools.

Fouser, T. J.↗
NASA NTRS1993

Reactive system verification case study: Fault-tolerant transputer communication

A reactive program is one which engages in an ongoing interaction with its environment. A system which is controlled by an embedded reactive program is called a reactive system. Examples of reactive systems are aircraft flight management systems, bank automatic teller machine (ATM) networks, airline reservation systems, and computer operating systems. Reactive systems are often naturally modeled (for logical design purposes) as a composition of autonomous processes which progress concurrently and which communicate to share information and/or to coordinate activities. Formal (i.e., mathematical) frameworks for system verification are tools used to increase the users' confidence that a system design satisfies its specification. A framework for reactive system verification includes formal languages for system modeling and for behavior specification and decision procedures and/or proof-systems for verifying that the system model satisfies the system specifications. Using the Ostroff framework for reactive system verification, an approach to achieving fault-tolerant communication between transputers was shown to be effective. The key components of the design, the decoupler processes, may be viewed as discrete-event-controllers introduced to constrain system behavior such that system specifications are satisfied. The Ostroff framework was also effective. The expressiveness of the modeling language permitted construction of a faithful model of the transputer network. The relevant specifications were readily expressed in the specification language. The set of decision procedures provided was adequate to verify the specifications of interest. The need for improved support for system behavior visualization is emphasized.

Crane, D. Francis↗
DOE OSTI2022

CrossSim Inference Manual v2.0

Neural networks are largely based on matrix computations. During forward inference, the most heavily used compute kernel is the matrix-vector multiplication (MVM): $W \vec{x} $. Inference is a first frontier for the deployment of next-generation hardware for neural network applications, as it is more readily deployed in edge devices, such as mobile devices or embedded processors with size, weight, and power constraints. Inference is also easier to implement in analog systems than training, which has more stringent device requirements. The main processing kernel used during inference is the MVM.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2021

Reduced-order modeling of advection-dominated systems with recurrent neural networks and convolutional autoencoders

A common strategy for the dimensionality reduction of nonlinear partial differential equations (PDEs) relies on the use of the proper orthogonal decomposition (POD) to identify a reduced subspace and the Galerkin projection for evolving dynamics in this reduced space. However, advection-dominated PDEs are represented poorly by this methodology since the process of truncation discards important interactions between higher-order modes during time evolution. In this study, we demonstrate that encoding using convolutional autoencoders (CAEs) followed by a reduced-space time evolution by recurrent neural networks overcomes this limitation effectively. We demonstrate that a truncated system of only two latent space dimensions can reproduce a sharp advecting shock profile for the viscous Burgers equation with very low viscosities, and a six-dimensional latent space can recreate the evolution of the inviscid shallow water equations. Additionally, the proposed framework is extended to a parametric reduced-order model by directly embedding parametric information into the latent space to detect trends in system evolution. Furthermore, our results show that these advection-dominated systems are more amenable to low-dimensional encoding and time evolution by a CAE and recurrent neural network combination than the POD-Galerkin technique.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2022

Scalable multiscale modeling of platelets with 100 million particles

Here, we developed the core components of the AI-aided multiple time stepping algorithm for multiscale modeling of cell dynamics. This algorithm was implemented and analyzed on two supercomputer architectures with an application of simulating the aggregation of 250 platelets, or 102 million particles. To scale on these computers with complex memory and network architectures with GPUs, we devised a biomechanics-informed task mapping scheme to optimize load imbalance, communications, and memory utilization. Our simulations, scaling well up to 192 nodes on a Summit-like supercomputer with a peak speed of 11 petaflops, achieved a rate of 423 μs/day which is 500 times faster than the conventional algorithm using static time step and this has enabled studies of record size blood clots at record spatial–temporal resolutions. Additionally, we discovered the sensitive dependence of the scalability and execution time on the methods of decomposition, CPU–GPU coupling, and task mapping.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2024

Speeding up and reducing memory usage for scientific machine learning via mixed precision

Scientific machine learning (SciML) has emerged as a versatile approach to address complex computational science and engineering problems. Within this field, physics-informed neural networks (PINNs) and deep operator networks (DeepONets) stand out as the leading techniques for solving partial differential equations by incorporating both physical equations and experimental data. However, training PINNs and DeepONets require significant computational resources, including long computational times and large amounts of memory. In search of computational efficiency, training neural networks using half precision (float16) rather than the conventional single (float32) or double (float64) precision has gained substantial interest, given the inherent benefits of reduced computational time and memory consumed. However, we find that float16 cannot be applied to SciML methods, because of gradient divergence at the start of training, weight updates going to zero, and the inability to converge to a local minima. To overcome these limitations, we explore mixed precision, which is an approach that combines the float16 and float32 numerical formats to reduce memory usage and increase computational speed. Our experiments showcase that mixed precision training not only substantially decreases training times and memory demands but also maintains model accuracy. Here, we also reinforce our empirical observations with a theoretical analysis. The research has broad implications for SciML in various computational applications.

97 MATHEMATICS AND COMPUTING↗
NASA NTRS1998

Experiments to Determine Whether Recursive Partitioning (CART) or an Artificial Neural Network Overcomes Theoretical Limitations of Cox Proportional Hazards Regression

New computationally intensive tools for medical survival analyses include recursive partitioning (also called CART) and artificial neural networks. A challenge that remains is to better understand the behavior of these techniques in effort to know when they will be effective tools. Theoretically they may overcome limitations of the traditional multivariable survival technique, the Cox proportional hazards regression model. Experiments were designed to test whether the new tools would, in practice, overcome these limitations. Two datasets in which theory suggests CART and the neural network should outperform the Cox model were selected. The first was a published leukemia dataset manipulated to have a strong interaction that CART should detect. The second was a published cirrhosis dataset with pronounced nonlinear effects that a neural network should fit. Repeated sampling of 50 training and testing subsets was applied to each technique. The concordance index C was calculated as a measure of predictive accuracy by each technique on the testing dataset. In the interaction dataset, CART outperformed Cox (P less than 0.05) with a C improvement of 0.1 (95% Cl, 0.08 to 0.12). In the nonlinear dataset, the neural network outperformed the Cox model (P less than 0.05), but by a very slight amount (0.015). As predicted by theory, CART and the neural network were able to overcome limitations of the Cox model. Experiments like these are important to increase our understanding of when one of these new techniques will outperform the standard Cox model. Further research is necessary to predict which technique will do best a priori and to assess the magnitude of superiority.

Kattan, Michael W.↗
NASA NTRS1991

Depth-size tradeoffs for neural computation

The tradeoffs between the depth (i.e., the time for parallel computation) and the size (i.e., the number of threshold gates) in neural networks are studied. The authors focus the study on the neural computations of symmetric Boolean functions and some arithmetic functions. It is shown that a significant reduction in the size is possible for symmetric functions and some arithmetic functions, at the expense of a small constant increase in depth. In the process, several neural networks which have the minimum size among all the known constructions have been developed. Results on implementing symmetric functions can be used to improve results about arbitrary Boolean functions. In particular, It is shown that any Boolean function can be computed in a depth-3 neural network with O(2n/2) threshold gates; it is also proven that a minimum number of threshold gates are required.

Siu, Kai-Yeung↗
NASA NTRS1994

Neural networks: Application to medical imaging

The research mission is the development of computer assisted diagnostic (CAD) methods for improved diagnosis of medical images including digital x-ray sensors and tomographic imaging modalities. The CAD algorithms include advanced methods for adaptive nonlinear filters for image noise suppression, hybrid wavelet methods for feature segmentation and enhancement, and high convergence neural networks for feature detection and VLSI implementation of neural networks for real time analysis. Other missions include (1) implementation of CAD methods on hospital based picture archiving computer systems (PACS) and information networks for central and remote diagnosis and (2) collaboration with defense and medical industry, NASA, and federal laboratories in the area of dual use technology conversion from defense or aerospace to medicine.

Clarke, Laurence P.↗
NASA NTRSDate not supplied

Implementation of a Hybrid Edge Node-Centroid Node Approach for the Generation of Reduced Thermal Models

Reduced thermal models are often required for delivery to organizations that manage models at the highest levels of assembly. The effort to generate/verify these models against their detailed counterparts is a challenge not yet conclusively solved. Higher level organizations often place a limit on the node count for delivered models, assuming smaller models result in faster computation. The burden of producing and verifying the accuracy of reduced models is primarily on the lower-level organizations, which consumes resources to produce these models. Limiting the number of allowable nodes may prevent users from taking full advantage of software capabilities that allow for faster generation of models, such as finite elements, which require more nodes than centroid based models. A methodology using Thermal Desktop was described in 2010 using: (1) finite elements and edge nodes for the conduction matrix, (2) centroid nodes for capacitance and radiative computations, and (3) the super network feature to produce a conduction matrix based only on the centroid nodes. While the methodology was clear, the implementation would have had to be done manually; with the inclusion of the OpenTD Application Programming Interface, this methodology can now be implemented programmatically and be a viable approach for the generation of reduced models. The approach was implemented at NASA-Goddard for the Capture-Containment-and-Return-System (CCRS) payload, resulting in the TCYEE tool. CCRS requires delivery with node limitations through the European Space Agency to its contractors. TCYEE generated the reduced model for delivery and the predictions compared favorably to the detailed model. Furthermore, TCYEE is being explored for potential use on the Roman Space Telescope (RST) for the generation of reduced models for delivery to the launch provider, which also has node limit requirements. This paper describes the methodology, implementation, and the performance of reduced models compared to their detailed counterparts.

Computer Programming and Software↗
DOE OSTI2026

What do physics-informed DeepONets learn? Understanding and improving training for scientific computing applications

Physics-informed deep operator networks (DeepONets) have emerged as a promising approach toward numerically approximating the solution of partial differential equations (PDEs). In this work, we aim to develop further understanding of what is being learned by physics-informed DeepONets by assessing the universality of the extracted basis functions and demonstrating their potential toward model reduction with spectral methods. Results provide clarity about measuring the performance of a physics-informed DeepONet through the decays of singular values and expansion coefficients. In addition, we propose a transfer learning approach for improving training for physics-informed DeepONets between parameters of the same PDE as well as across different, but related, PDEs where these models struggle to train well. This approach results in significant error reduction and learned basis functions that are more effective in representing the solution of a PDE.

Deep operator networks↗
NASA NTRS1991

Towards practical control design using neural computation

The objective is to develop neural network based control design techniques which address the issue of performance/control effort tradeoff. Additionally, the control design needs to address the important issue if achieving adequate performance in the presence of actuator nonlinearities such as position and rate limits. These issues are discussed using the example of aircraft flight control. Given a set of pilot input commands, a feedforward net is trained to control the vehicle within the constraints imposed by the actuators. This is achieved by minimizing an objective function which is the sum of the tracking errors, control input rates and control input deflections. A tradeoff between tracking performance and control smoothness is obtained by varying, adaptively, the weights of the objective function. The neurocontroller performance is evaluated in the presence of actuator dynamics using a simulation of the vehicle. Appropriate selection of the different weights in the objective function resulted in the good tracking of the pilot commands and smooth neurocontrol. An extension of the neurocontroller design approach is proposed to enhance its practicality.

Troudet, Terry↗
DOE OSTI2023

A Systematic Study to Determine 5G Baseline Performance for Scientific Computing

The fifth-generation (5G) cellular networks envisions achieving higher data rates, improved connectivity, reduced latency, and better quality of service (QoS) than the fourthgeneration (4G) cellular networks. Such improved performance can be utilized to address the challenges in applications such as electricity generation in power systems. The traditional power grids responsible for electricity generation suffer from drawbacks such as life-threatening blackout crises, and energy storage proliferation as they are not robust to extreme climatic conditions. A recent study proposed the idea of extending the capabilities of advanced wireless technologies such as the current 5G to develop a robust, energy-efficient, and secure smart grids. However there are two main challenges associated with the integration of power systems and wireless technologies. First, it is imperative to understand the architecture and the enabling technologies of 5G to ensure that the performance requirements of the smart grids are met. Second, an end-to-end testbed is required to determine if the performance requirements are met by estimating the 5G characteristics such as latency, and throughput. Our proposed alleviates the aforementioned concerns in the following manner. To begin with, a systematic study of the 5G architecture including both the StandAlone (SA) and Non-Standalone (NSA) operations is presented. Furthermore, a detailed survey of the possible 5G enabling technologies is elicited. In addition to these, an end-toend testbed that can estimate the 5G characteristics is explained in detail with appropriate preliminary results.

5G, 5G Communication↗
NASA NTRS1986

The new Langley Research Center advanced real-time simulation (ARTS) system

Based on a survey of current local area network technology with special attention paid to high bandwidth and very low transport delay requirements, NASA's Langley Research Center designed a new simulation subsystem using the computer automated measurement and control (CAMAC) network. This required significant modifications to the standard CAMAC system and development of a network switch, a clocking system, new conversion equipment, new consoles, supporting software, etc. This system is referred to as the advanced real-time simulation (ARTS) system. It is presently being built at LaRC. This paper provides a functional and physical description of the hardware and a functional description of the software. The requirements which drove the design are presented as well as present performance figures and status.

Crawford, D. J.↗
NASA NTRS1990

Analyzing Distributed Processing For Electric Utilities

Distributed Processing Trade-Off Model for Electric Utility Operation computer program based upon study performed at California Institute of Technology for NASA's Jet Propulsion Laboratory. Study presented technique addressing question of tradeoffs between expanding communications network or expanding capacity of distributed computers in energy-management systems (EMS) of electric utility. Gives EMS planners macroscopic tool for evaluation of architectures of distributed-processing systems and major technical and economic tradeoffs as well as interactions within systems.

Klein, Stanley A.↗