Search NASA⌕ Search

SEARCH · Search NASA

Results for “Network parameter error”

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 19 records

Multi-area parameter error identification for large power systems

Power grid model parameters may contain errors due to various reasons. Detecting and correcting parameter errors typically requires significant computational effort due to the size and complexity of the parameter database. While the normalized Lagrange multiplier (NLM) method can effectively detect, identify and correct parameter errors, its computational burden could rapidly grow with increasing system size. This paper addresses this issue by proposing a multi-area parameter error identification method. Each area has its own outlier detection tool for detecting the incorrect parameters and measurements within the area. On the other hand, due to the reduced redundancy at area boundaries, parameter errors on branches incident to boundary buses may not be detected. Such errors are subsequently detected by a coordination level estimator completing the system-wide parameter detection procedure. In conclusion, performance of the developed method is demonstrated using the IEEE 118-bus and 2000-bus Texas synthetic systems.

24 POWER TRANSMISSION AND DISTRIBUTION↗

Deep material network via a quilting strategy: visualization for explainability and recursive training for improved accuracy

Recent developments integrating micromechanics and neural networks offer promising paths for rapid predictions of the response of heterogeneous materials with similar accuracy as direct numerical simulations. The deep material network is one such approaches, featuring a multi-layer network and micromechanics building blocks trained on anisotropic linear elastic properties. Once trained, the network acts as a reduced-order model, which can extrapolate the material’s behavior to more general constitutive laws, including nonlinear behaviors, without the need to be retrained. However, current training methods initialize network parameters randomly, incurring inevitable training and calibration errors. Here, we introduce a way to visualize the network parameters as an analogous unit cell and use this visualization to “quilt” patches of shallower networks to initialize deeper networks for a recursive training strategy. The result is an improvement in the accuracy and calibration performance of the network and an intuitive visual representation of the network for better explainability.

97 MATHEMATICS AND COMPUTING↗

Learning quantum computers' errors using interpretable neural networks

Learning and reducing the errors and noise in quantum computing systems is necessary for achieving quantum computation’s promise. However, rapid advances in experimental quantum computing are making this task increasingly difficult, because state-of-the-art systems now contain hundreds of qubits and many characterization techniques are hard to apply at this scale. Furthermore, complex kinds of errors in these systems, such as crosstalk and non-Markovian effects, must be understood and decreased, but these errors are challenging to study with most existing methods. In this project, we explored using neural networks for scalable characterization of complex errors in quantum computers. We proposed and demonstrated characterizing a quantum computer’s errors with neural networks that have interpretable parameters corresponding to the rates of different kinds of errors, within a sparse Lindbladian parameterization for errors. To enable scaling to many qubit systems, these networks then predict how these errors combine within quantum circuits and impact their outcomes using an efficient approximations. We demonstrated these networks ability to learn coherent crosstalk errors and context-dependent errors in a simulated 4-qubit system.

97 MATHEMATICS AND COMPUTING↗

Commissioning of SRF Cavity Analyzer System

Research and development in Superconducting Radio Frequency (SRF) cavities manufacturing, surface treatment, preparation, and doping techniques for the improvement of cavities performance necessitate reliable cavities’ intrinsic quality factor measurement with accurate and repeatable results. Fermilab Vertical Test Stand facility (VTS) performs ~270 cavities tests per year, including R&D cavities as well as production cavities at each step of their preparation for the placement in accelerator’s cryomodules. Current Fermilab VTS testing system is based on the scalar approach of power measurement, which neglects potential error factors of the RF networks with distributed parameters, hence the assessed measurement errors of such a system might be slightly underestimated. Alternative approach, utilizing a vector measurement (amplitude and phase) of RF signals in the setup with a cavity as a device under test, promises an enhanced accuracy and results traceability. New cavity an alyser system based on this approach (Vector VTS) was developed by D. Frolov et al. and currently undergoes a commissioning process. Basic underlying principles of the system will be presented with the methods of quality factor extraction from S-parameters measurement of multiport cavities and brief error analysis will be discussed. The system architecture will be shown with main components parameters and ensuing testing capabilities. Multiple tests results will be discussed with a cross-check comparison of the results obtained with a conventional testing method. System advantages such as automated calibration process and additional features as a built- in microphonics measurement capability will be presented.

Netepenko, A.↗

Inversion of surface parameters using fast learning neural networks

A neural network approach to the inversion of surface scattering parameters is presented. Simulated data sets based on a surface scattering model are used so that the data may be viewed as taken from a completely known randomly rough surface. The fast learning (FL) neural network and a multilayer perceptron (MLP) trained with backpropagation learning (BP network) are tested on the simulated backscattering data. The RMS error of training the FL network is found to be less than one half the error of the BP network while requiring one to two orders of magnitude less CPU time. When applied to inversion of parameters from a statistically rough surface, the FL method is successful at recovering the surface permittivity, the surface correlation length, and the RMS surface height in less time and with less error than the BP network. Further applications of the FL neural network to the inversion of parameters from backscatter measurements of an inhomogeneous layer above a half space are shown.

Dawson, M. S.↗

Common Electric Power Transmission System Model JSON Schema Specification

The Common Electric Power Transmission System Model (CTM) is an intuitive, extensible, language-agnostic, and error-resistant specification of electric power network components parameter names and units, and relation between components, intended for use by the research community developing new computational methods for power systems operations and simulation. Power system datasets following the CTM specification can be read as dictionaries and manipulated in that form in most programming languages (e.g., Python, Julia, C++). This standard data structure in CTM makes it easy to work in multiple power systems domains (e.g., economic operation, reliability assessment, electricity markets, stability assessment, etc.) without requiring conversions between use-case-specific file formats with information loss in the process. This repository specifies CTM as a JSON Schema, provides documentation, derivate (code-generated) implementations of CTM, and example data and usage of the schema for important use cases.

Aravena Solis, Ignacio↗

User-Friendly Interface Developed for a Web-Based Service for SpaceCAL Emulations

A team at the NASA Glenn Research Center is developing a Space Communications Architecture Laboratory (SpaceCAL) for protocol development activities for coordinated satellite missions. SpaceCAL will provide a multiuser, distributed system to emulate space-based Internet architectures, backbone networks, formation clusters, and constellations. As part of a new effort in 2003, building blocks are being defined for an open distributed system to make the satellite emulation test bed accessible through an Internet connection. The first step in creating a Web-based service to control the emulation remotely is providing a user-friendly interface for encoding the data into a well-formed and complete Extensible Markup Language (XML) document. XML provides coding that allows data to be transferred between dissimilar systems. Scenario specifications include control parameters, network routes, interface bandwidths, delay, and bit error rate. Specifications for all satellite, instruments, and ground stations in a given scenario are also included in the XML document. For the SpaceCAL emulation, the XML document can be created using XForms, a Webbased forms language for data collection. Contrary to older forms technology, the interactive user interface makes the science prevalent, not the data representation. Required versus optional input fields, default values, automatic calculations, data validation, and reuse will help researchers quickly and accurately define missions. XForms can apply any XML schema defined for the test mission to validate data before forwarding it to the emulation facility. New instrument definitions, facilities, and mission types can be added to the existing schema. The first prototype user interface incorporates components for interactive input and form processing. Internet address, data rate, and the location of the facility are implemented with basic form controls with default values provided for convenience and efficiency using basic XForms operations. Because different emulation scenarios will vary widely in their component structure, more complex operations are used to add and delete facilities.

Liszka, Kathy J.↗

Machine Learning Methods for Estimating Propeller Source Noise Spheres

In this work, several neural network function approximations are compared for inter- polating, storing, and sampling acoustic source spheres with applications to propeller noise estimation. These methods are compared using an acoustic model of the three bladed GL-10 propeller at different flight conditions, with training data generated using NASA’s ANOPP-PAS module. The source spheres used to train the networks capture the tonal propeller noise due to both the blade thickness and loading. This tonal noise prediction method allows the vehicle noise to be estimated for auralization and acoustic control. Three radial basis function neural network architectures are compared in this work. The first two networks directly estimate the parameters of the source sphere at different flight conditions but differ in the number of layers used. The third network estimates the parameters of the source sphere using a weighted combination of spherical basis functions. These networks are trained on numerically generated source spheres, with operating points given in terms of the propeller rotation rate, freestream speed, and propeller angle of attack. The performance of the neural network is determined using a validation dataset of withheld data points. This performance is quantified in terms of the approximation error, training time, and sample time. The third network, which estimates the weights of the spherical basis functions, performs the best in both average and maximum approximation errors in all cases. This network’s worst case performance is 5.6 % relative dif- ference of a model parameter associated with acoustic pressure. The direct estimation network with a single layer has the worst approximation error in all cases. Additionally, the spherically defined network has the slowest sample time at 0.05 seconds per thousand points. Both direct estimation methods produce a thousand sample points in approximately 0.001 seconds.

Acoustics↗

Fast temporal neural learning using teacher forcing

A neural network is trained to output a time dependent target vector defined over a predetermined time interval in response to a time dependent input vector defined over the same time interval by applying corresponding elements of the error vector, or difference between the target vector and the actual neuron output vector, to the inputs of corresponding output neurons of the network as corrective feedback. This feedback decreases the error and quickens the learning process, so that a much smaller number of training cycles are required to complete the learning process. A conventional gradient descent algorithm is employed to update the neural network parameters at the end of the predetermined time interval. The foregoing process is repeated in repetitive cycles until the actual output vector corresponds to the target vector. In the preferred embodiment, as the overall error of the neural network output decreasing during successive training cycles, the portion of the error fed back to the output neurons is decreased accordingly, allowing the network to learn with greater freedom from teacher forcing as the network parameters converge to their optimum values. The invention may also be used to train a neural network with stationary training and target vectors.

Toomarian, Nikzad↗

Fast temporal neural learning using teacher forcing

A neural network is trained to output a time dependent target vector defined over a predetermined time interval in response to a time dependent input vector defined over the same time interval by applying corresponding elements of the error vector, or difference between the target vector and the actual neuron output vector, to the inputs of corresponding output neurons of the network as corrective feedback. This feedback decreases the error and quickens the learning process, so that a much smaller number of training cycles are required to complete the learning process. A conventional gradient descent algorithm is employed to update the neural network parameters at the end of the predetermined time interval. The foregoing process is repeated in repetitive cycles until the actual output vector corresponds to the target vector. In the preferred embodiment, as the overall error of the neural network output decreasing during successive training cycles, the portion of the error fed back to the output neurons is decreased accordingly, allowing the network to learn with greater freedom from teacher forcing as the network parameters converge to their optimum values. The invention may also be used to train a neural network with stationary training and target vectors.

Toomarian, Nikzad↗

Identification and control of plasma vertical position using neural network in Damavand tokamak

In this work, a nonlinear model is introduced to determine the vertical position of the plasma column in Damavand tokamak. Using this model as a simulator, a nonlinear neural network controller has been designed. In the first stage, the electronic drive and sensory circuits of Damavand tokamak are modified. These circuits can control the vertical position of the plasma column inside the vacuum vessel. Since the vertical position of plasma is an unstable parameter, a direct closed loop system identification algorithm is performed. In the second stage, a nonlinear model is identified for plasma vertical position, based on the multilayer perceptron (MLP) neural network (NN) structure. Estimation of simulator parameters has been performed by back-propagation error algorithm using Levenberg-Marquardt gradient descent optimization technique. The model is verified through simulation of the whole closed loop system using both simulator and actual plant in similar conditions. As the final stage, a MLP neural network controller is designed for simulator model. In the last step, online training is performed to tune the controller parameters. Simulation results justify using of the NN controller for the actual plant.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Improving the Representation of Land Surface Processes Using the Data Assimilation Research Testbed (DART)

The land surface is a critical part of the earth system as processes related to water, carbon, energy and nitrogen cycling have important implications for climate forcing, air quality, water availability and seasonal atmospheric forecasting. Despite advances in land surface modeling, land surface model performance is often limited because of errors related to initial and boundary conditions, model structure, and parameters. Data assimilation (DA) techniques combined with an expanding network of earth system observations present an opportunity to reduce these errors and improve simulations. Here, we emphasize the implementation of tools and approaches to overcome challenges related to land DA to constrain carbon and water cycling. In particular, we discuss the implementation of adaptive inflation to modify ensemble spread in response to time-varying networks of gridded observations. We also discuss methods to generate ensemble spread through boundary condition (meteorology) forcing that can be applied to site-level applications. Next, we describe the application of vertical localization upon surface soil moisture observations, and forward operators specifically designed for the assimilation of snow and solar-induced fluorescence observations. Finally, we discuss the potential benefit of a quantile conserving filter used to update bounded quantities (state or parameter values).

Brett Raczka↗

FedCSpc: A Cross-Silo Federated Learning System With Error-Bounded Lossy Parameter Compression

Cross-Silo federated learning is widely used for scaling deep neural network (DNN) training over data silos from different locations worldwide while guaranteeing data privacy. Communication has been identified as the main bottleneck when training large-scale models due to large-volume model parameters and gradient transmission across public networks with limited bandwidth. Most previous works focus on gradient compression, while limited work tries to compress parameters that can not be ignored and extremely affect communication performance during the training. Here, to bridge this gap, we propose FedCSpc: an efficient cross-silo federated learning system with an XAI-driven adaptive parameter compression strategy for large-scale model training. Our work substantially differs from existing gradient compression techniques due to the distinct data features of gradient and parameter. The key contributions of this paper are fourfold. (1) Our designed FedCSpc proposes to compress the parameter during the training using the state-of-the-art error-bounded lossy compressor – SZ3. (2) We develop an adaptive compression error bound adjustment algorithm to guarantee the model accuracy effectively. (3) We exploit an efficient approach to utilize the idle CPU resources of clients to compress the parameters. (4) We perform a comprehensive evaluation with a wide range of models and benchmarks on a GPU cluster with 65 GPUs. Results show that FedCSpc can achieve the same model accuracy as FedAvg while reducing the data volume of parameters and gradients in communication by up to 7.39× and 288×, respectively. With 32 clients on a 4 Gb size model, FedCSpc significantly outperforms FedAvg in wall-clock time in the emulated WAN environment (at the bandwidth of 1 Gbps or lower without loss of generality).

SZ3↗

Transforming microseismic clouds into near real-time visualization of the growing hydraulic fracture

SUMMARY Microseismic observations during unconventional reservoir stimulation are typically seen as a proxy for clusters of hydraulic fractures and the extent of the stimulated reservoir. Such straightforward interpretation is often misleading and fails to provide a physically reasonable image of the fracturing process. This paper demonstrates the application of a physics-based machine learning algorithm which enables a rapid and accurate fracture mapping from the microseismic data. Our training and validation data set relies on a history-matched geomechanical modelling workflow implemented in GEOS software for the Hydraulic Fracturing Test Site 1 (HFTS-1) project. For this study we augmented the simulated fracture growth through geostatistical modelling of induced seismicity, so that the synthetic microseismic catalogue matches the main statistical properties of the field observations. We formulated the problem of mapping the actual fracture in the clutter of events to parallel common video segmentation workflows: several past video frames (microseismic density snapshots) are passed through a deep convolutional network to classify whether a given voxel is associated with a fracture or intact rock. We found that for accurate fracture mapping, the network’s input and architecture must be augmented to incorporate the fluid injection parameters (pressure, rate, concentration of proppant, and location of the perforation within the cluster). The error rate for the network reached as little as 10 per cent of the fracture area, while a conventional microseismic interpretation approach yielded ∼300 per cent. Our approach also yields must faster predictions than conventional methods (minutes instead of weeks), and could enable engineers to make rapid decisions regarding engineering parameters (pumping rate, viscosity) in real time during stimulation.

58 GEOSCIENCES↗

Improving the Representation of Land Surface Processes using the Data Assimilation Research Testbed (DART)

The land surface is a critical part of the earth system as processes related to water, carbon, energy and nitrogen cycling have important implications for climate forcing, air quality, water availability and seasonal atmospheric forecasting. Despite advances in land surface modeling, land surface model performance is often limited because of errors related to initial and boundary conditions, model structure, and parameters. Data assimilation (DA) techniques combined with an expanding network of earth system observations present an opportunity to reduce these errors and improve simulations. Here we apply an Ensemble Kalman Filter DA system as part of the Data Assimilation Research Testbed to a variety of land surface simulations. First, we describe the use of remotely sensed biomass observations to provide improved simulations of plant phenology, carbon and water cycling for regions highly sensitive to climate change (Western US, China, and Arctic). We discuss approaches to account for systemic biases between models and observations, including the use of spatially-varying adaptive ensemble inflation as an alternative approach to re-scaling soil moisture observations. Finally, we discuss a strategy to incorporate complementary observations (snow water equivalent, solar-induced fluorescence) to better constrain the representation of carbon and water cycling across complex terrain.

DART↗

SU-E-T-51: Bayesian Network Models for Radiotherapy Error Detection

Purpose: To develop a probabilistic model of radiotherapy plans using Bayesian networks that will detect potential errors in radiation delivery. Methods: Semi-structured interviews with medical physicists and other domain experts were employed to generate a set of layered nodes and arcs forming a Bayesian Network (BN) which encapsulates relevant radiotherapy concepts and their associated interdependencies. Concepts in the final network were limited to those whose parameters are represented in the institutional database at a level significant enough to develop mathematical distributions. The concept-relation knowledge base was constructed using the Web Ontology Language (OWL) and translated into Hugin Expert Bayes Network files via the the RHugin package in the R statistical programming language. A subset of de-identified data derived from a Mosaiq relational database representing 1937 unique prescription cases was processed and pre-screened for errors and then used by the Hugin implementation of the Estimation-Maximization (EM) algorithm for machine learning all parameter distributions. Individual networks were generated for each of several commonly treated anatomic regions identified by ICD-9 neoplasm categories including lung, brain, lymphoma, and female breast. Results: The resulting Bayesian networks represent a large part of the probabilistic knowledge inherent in treatment planning. By populating the networks entirely with data captured from a clinical oncology information management system over the course of several years of normal practice, we were able to create accurate probability tables with no additional time spent by experts or clinicians. These probabilistic descriptions of the treatment planning allow one to check if a treatment plan is within the normal scope of practice, given some initial set of clinical evidence and thereby detect for potential outliers to be flagged for further investigation. Conclusion: The networks developed here support the use of probabilistic models into clinical chart checking for improved detection of potential errors in RT plans.

60 APPLIED LIFE SCIENCES↗