Search NASA⌕ Search

SEARCH · Search NASA

Results for “Optimization methods”

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 649 records · Page 36
All sourcesNASA NTRSDOE OSTI
NASA NTRS1994

Optimal Low Thrust Trajectories Using Differential Inclusion Concepts

A direct optimization method based on differential inclusion concepts has been developed and used to compute low thrust trajectories. This new formulation removes explicit control dependence from the problem statement thereby reducing the dimension of the parameter space of the resulting nonlinear programming problem. A simple example of a two-dimensional gravity-free trajectory involving a maximum velocity transfer to a rectilinear path is discussed. Three interplanetary trajectory examples, an Earth-Mars constant specific impulse transfer, and Earth-Jupiter constant specific impulse transfer, and an Earth-Venus-Mars variable specific impulse gravity assist, are also included. An analysis of the technique's performance is provided.

low↗
NASA NTRSDate not supplied

Use of Design of Experiments in Determining Neural Network Architectures for Loss of Control Detection

Abstract—We describe empirical methods for selecting a neural network architecture to implement belief state inference on generic commercial transport aircraft. We highlight a case study on the planning, execution, and analysis of a set of experiments to determine the configurations of a conditional variational autoencoder (CVAE). Our main contribution is the application of a structured method that can be used for machine learning in many aerospace applications. This method optimizes the structure and training parameters of a neural network for belief state inference, using Design of Experiments (DOE) statistical methodologies. The motivation for this specific DOE analysis was to identify the appropriate hyperparameters for measuring the CVAE reconstruction probability and latent space, such that the measurements can be used to infer qualitative state changes for the aircraft. We demonstrate that this process yields information about a trained neural network’s utility for this specific application, along with a quantifiable range of certainty. We execute 84 experiments using loss-of-control flight maneuver data from the NASA T-2 aircraft, demonstrating that this empirical process allows us to construct cheap and simple models with specific attributes amenable to belief state inference in aerospace applications.

neural networks↗
NASA NTRSDate not supplied

An Automated Refinement Process for Particle Trajectory Methods in GlennICE

Computational methods for ice accretion can simulate the impact of water drops and ice crystals on an aircraft surface in a Lagrangian reference frame or in the Eulerian reference frame. In the Eulerian reference frame, particles are considered a continuous fluid while in the Lagrangian frame individual particle trajectories are calculated. Methods that use the Eulerian reference frame are typically easier to develop as established modules used for continuum mechanics can be leveraged. The Eulerian systems can also be faster since the user does not have to simulate millions of particles in order to achieve good results. It is imperative therefore that a Lagrangian method optimize the release points of trajectories such that accurate solutions can be obtained while minimizing as much as possible the number of trajectories computed. This paper will present a methodology for this refinement process and demonstrate its effectiveness on sample three dimensional test cases.

William B Wright↗
NASA NTRSDate not supplied

An Automated Refinement Process for Particle Trajectory Methods in GlennICE

Computational methods for ice accretion can simulate the impact of water drops and ice crystals on an aircraft surface in a Lagrangian reference frame or in the Eulerian reference frame. In the Eulerian reference frame, particles are considered a continuous fluid while in the Lagrangian frame individual particle trajectories are calculated. Methods that use the Eulerian reference frame are typically easier to develop as established modules used for continuum mechanics can be leveraged. The Eulerian systems can also be faster since the user does not have to simulate millions of particles in order to achieve good results. It is imperative therefore that a Lagrangian method optimize the release points of trajectories such that accurate solutions can be obtained while minimizing as much as possible the number of trajectories computed. This paper will present a methodology for this refinement process and demonstrate its effectiveness on sample three dimensional test cases.

William Wright↗
NASA NTRS2020

Tuning a variational autoencoder for data accountability problem in the Mars Science Laboratory ground data system

The Mars Curiosity rover is frequently sending back engineering and science data that goes through a pipeline of systems before reaching its final destination at the mission operations center making it prone to volume loss and data corruption. A ground data system analysis (GDSA) team is charged with the monitoring of this flow of information and the detection of anomalies in that data in order to request a re-transmission when necessary. This work presents ∆-MADS, a derivative-free optimization method applied for tuning the architecture and hyperparameters of a variational autoencoder trained to detect the data with missing patches in order to assist the GDSA team in their mission.

Lakhmiri, Dounia↗
NASA NTRS2020

Optimal Low Thrust Orbit Transfers for Space Telescope Refueling at SEL2

The James Webb Space Telescope (JWST), a ten billion-dollar infrared telescope with a 6.5m primary mirror to be launched in 2021, is designed to operate in a Halo orbit around the second Sun-Earth Lagrange point (SEL2) for five to ten years. At that point fuel for station keeping and attitude maneuvers will run out. Refueling missions to JWST, as well as to similar space telescope missions proposed for SEL2, could greatly enhance the “science-per-dollar” value and promote a more sustainable use of space assets. In this paper, we present a novel approach to designing fuel optimal trajectories that will allow the refueling spacecraft to arrive at the SEL2 Halo orbit with maximum final mass (i.e. fuel payload). The low thrust optimal control problem is formulated using an indirect optimization method, leading to a two-point boundary value problem with a bang-bang control structure. We make use of a hyperbolic tangent smoothing technique for performing continuation on the thrust magnitude to reduce the sharpness of the control switches in early iterations and, thus, promote convergence. The problem is posed and solved in the circular restricted three-body problem. In this dynamical system, invariant manifolds exist that can be utilized to reduce fuel consumption. The here presented methodology to this challenging and important problem in astrodynamics demonstrates a significant potential for low-cost refueling mission design.

Woollands, Robyn↗
NASA NTRS2022

P- and L-Band Retrieval of Subsurface Soil Moisture and Temperature Profiles as First-Order Polynomial Function

This paper demonstrates the potential use of P and L band passive measurements to determine root zone soil moisture (SM) and soil temperature(ST). SM and ST data have been taken as a function of depth during the NASA GSFC PLEX19 experiment in the summer of 2019 at Beltsville, MD, USA. Using these data, a coherent model has been used to compute H and V brightness temperatures at frequencies of 0.8 and 1.4 GHz with an observation angle of 35 degrees. These synthetic brightness data are then used to estimate the SM and ST profiles which are represented by linear polynomials. The inversion problem is formulated as a least square problem that is solved by a global optimization method known as the Adaptive Simulated Annealing(ASA) method. Four inversion examples having different SM and ST profiles are presented. Selected results show that the standard deviation between the retrieved and measured data is less than 0.077 cm3/cm3 for SM, and 2.245 °C for ST.

Ming Li↗
NASA NTRS2021

Error Propagation in Microwave Soil Moisture and Vegetation Optical Depth Retrievals

Satellite soil moisture and vegetation optical depth (VOD; related to the total vegetation water mass per unit area) are increasingly being used to study water relations in the soil-plant continuum across the globe. However, soil moisture and VOD are typically jointly estimated, where errors in the optimization approach can cause compensation between both variables and confound such studies. It is thus critical to quantify how satellite microwave measurement errors propagate into soil moisture and VOD. Such a study is especially important for VOD given limited investigations of whether VOD reflects in-situ plant physiology. Furthermore, despite new approaches that constrain (or regularize) VOD dynamics to reduce soil moisture errors, there is limited study of whether regularization reduces VOD errors without obscuring true vegetation temporal dynamics. Here, we find that, across the globe, VOD is less robust to measurement error (more difficult for optimization methods to find the true solution) than soil moisture in their joint estimation. However, a moderate degree of regularization (via time-constrained VOD) reduces errors in VOD to a greater degree than soil moisture and reduces spurious soil moisture-VOD coupling. Furthermore, despite constraining VOD time dynamics, regularized VOD variations on sub-weekly scales are both closer to simulated true VOD time series and have global VOD post-rainfall responses with reduced error signatures compared to VOD retrievals without regularization. Ultimately, we recommend moderately regularized VOD for use in large scale studies of soil-plant water relations because it suppresses noise and spurious soil moisture-VOD coupling without removing the physical signal.

Vegetation optical depth↗
NASA NTRS2024

Optimized bandpasses for the Habitable Worlds Observatory's exoEarth survey

A primary scientific goal of the future Habitable Worlds Observatory will be the direct detection and characterization of Earth-like planets. Estimates of the exoplanet yields for this concept will help guide mission design through detailed trade studies. It is therefore critical that yield estimation codes optimally adapt observations to the mission’s performance parameters to ensure accurate trade studies. To aid in this, we implement wavelength optimization in yield calculations for the first time, allowing the yield code to determine the ideal detection and characterization bandpasses. We use this capability to confirm the observational wavelength assumptions made for the large UV/Optical/IR surveyor, design B (LUVOIR-B) study, namely that the optimum detection wavelength is 500 nm for the majority of targets and the optimum wavelength to detect water is near 1000 nm, given LUVOIR-B’s assumed instrument performance parameters. We show that including the wavelength-dependent albedo of an Earth twin as a prior provides no significant benefit to the yields of exoEarth candidates and caution against tuning observations to modern Earth twins. We also show that coronagraphs whose inner working angles are similar to step functions may benefit from wavelength optimization and demonstrate how wavelength-dependent instrument performance can impact the optimum wavelengths for detection and characterization. The optimization methods we implement automate wavelength selection and remove uncertainties regarding these choices, helping to adapt the observations to the instrument’s performance parameters.

telescopes↗
NASA NTRSDate not supplied

Developing a Genetic Variant Calling Pipeline for Quantifying the Complex Mutagenic Load Accumulated in BioNutrients-1 Production Pack Samples

Microorganisms hold great promise for on demand production of labile nutrients and pharmaceuticals as well recycling and in situ resource utilization. The utilization of microorganisms for such tasks on space missions is hindered by the limited data on how microbes respond to spaceflight. For example, the genetic stability of microorganisms, and the genomic engineered traits added to deliver desired functions, over long-term storage in the spacecraft environment is poorly understood. The BioNutrients-1 (BN-1) mission conducted a 5-year study of desiccated storage in Low Earth Orbit (LEO) to evaluate the suitability of eight synthetic biology chassis organisms for long-duration space missions. We are employing high-depth, whole genome sequencing (WGS) to determine the mutagenic load that accumulated during long-term storage. Mutation analysis pipelines are well established for homogenous culture grown from a single colony, but the mutational landscape of the BN-1 samples present a unique analysis challenge, as every cell in the BN-1 samples had a unique genetic journey of DNA damage and repair. Consequently, sequence variants are expected at low allele frequency within samples. To address this genetic complexity, we apply two distinct computational approaches to identify mutations in pre-existing WGS data collected from populations of Chlamydomonas reinhardtii that were exposed to UV mutagenesis and growth in LEO. For reference genome free mutation detection, we utilized DiscoSNP++, which is a de Bruijn graph approach. For reference genome-based mutation detection we utilize GATK for Microbes, which is a Bayesian probabilistic approach. We will benchmark these approaches against the mutations originally identified using CRISP, a method optimized for pooled samples. Ultimately, quantifying the mutation load imposed by storage or growth on the ISS will help identify chassis organisms with both high levels of genome stability and viability, which are desirable traits for implementation of bioproduction in long-duration missions.

SNP↗
DOE OSTI2025

Enhancing The Thermal Resistivity of Rigid Polyisocyanurate Foam Insulation

The development of rigid polyurethane foam insulation has garnered considerable attention because of its promising applications in the buildings and construction industry. Its low thermal conductivity makes it an attractive choice for improving energy performance in buildings. Current Rigid Polyurethane foams have thermal resistivity (R-value/in.) 5.5 to 6.5 h.ft2F/BTU/in.· that could be further improved by diminishing the heat transfer through the foam matrix. Nevertheless, minimizing both conduction (through gas and solid) and radiation simultaneously in porous solids is a significant challenge due to the trade-off between these two mechanisms. This study focuses on improving the R-value of the insulation foams via several strategies: such as type and the amount of the surfactants, blowing agent content and precooling and premixing polyol mixture. These methods optimized thermal properties of the PIR foams, achieving R/in. as high as 8.3. This excellent R/in. is anticipated to be a critical factor in significantly advancing the thermal insulation performance of rigid polyurethane cellular foams, thereby enhancing their efficacy in energy-efficient building applications.

Wanasinghe Mudiyanselage Pahala Gedara, Shiwanka V↗
DOE OSTI2025

A Brief Survey on High Performance Computing Systems Power Management

This paper provides a survey of software-based power management techniques in High Performance Computing (HPC) systems. Seven existing power management and monitoring tools and frameworks are discussed. These are: Variorum, dynamic energy-performance optimizer (DEPO), Powersched, Bull Dynamic Power Optimizer (BDPO), Energy Aware Runtime (EAR), Global Extensible Open Power Manager (GEOPM), and PoLiMEr. Each of these tools is evaluated based on hardware abstraction, optimization methods, usability, and experimental validation. This survey highlights the diversity of approaches in managing energy efficiency, from vendor-neutral APIs to algorithm-driven power capping, and dynamic frequency adjustments. Given that energy requirements for large computational systems is increasing quickly, the importance of integrating these tools into existing HPC environments and the need for further research in this rapidly evolving field is also discussed.

97 - MATHEMATICS AND COMPUTING↗
NASA NTRS2009

Robust Optimal Adaptive Control Method with Large Adaptive Gain

In the presence of large uncertainties, a control system needs to be able to adapt rapidly to regain performance. Fast adaptation is referred to the implementation of adaptive control with a large adaptive gain to reduce the tracking error rapidly. However, a large adaptive gain can lead to high-frequency oscillations which can adversely affect robustness of an adaptive control law. A new adaptive control modification is presented that can achieve robust adaptation with a large adaptive gain without incurring high-frequency oscillations as with the standard model-reference adaptive control. The modification is based on the minimization of the Y2 norm of the tracking error, which is formulated as an optimal control problem. The optimality condition is used to derive the modification using the gradient method. The optimal control modification results in a stable adaptation and allows a large adaptive gain to be used for better tracking while providing sufficient stability robustness. Simulations were conducted for a damaged generic transport aircraft with both standard adaptive control and the adaptive optimal control modification technique. The results demonstrate the effectiveness of the proposed modification in tracking a reference model while maintaining a sufficient time delay margin.

Nguyen, Nhan T.↗
NASA NTRS1991

A general weight matrix formulation using optimal control

Classical methods from optimal control theory are used in deriving general forms for neural network weights. The network learning or application task is encoded in a performance index of a general structure. Consequently, different instances of this performance index lead to special cases of weight rules, including some well-known forms. Comparisons are made with the outer product rule, spectral methods, and recurrent back-propagation. Simulation results and comparisons are presented.

Farotimi, Oluseyi↗
NASA NTRS1997

Gradient-Based Aerodynamic Shape Optimization Using ADI Method for Large-Scale Problems

A gradient-based shape optimization methodology, that is intended for practical three-dimensional aerodynamic applications, has been developed. It is based on the quasi-analytical sensitivities. The flow analysis is rendered by a fully implicit, finite volume formulation of the Euler equations.The aerodynamic sensitivity equation is solved using the alternating-direction-implicit (ADI) algorithm for memory efficiency. A flexible wing geometry model, that is based on surface parameterization and platform schedules, is utilized. The present methodology and its components have been tested via several comparisons. Initially, the flow analysis for for a wing is compared with those obtained using an unfactored, preconditioned conjugate gradient approach (PCG), and an extensively validated CFD code. Then, the sensitivities computed with the present method have been compared with those obtained using the finite-difference and the PCG approaches. Effects of grid refinement and convergence tolerance on the analysis and shape optimization have been explored. Finally the new procedure has been demonstrated in the design of a cranked arrow wing at Mach 2.4. Despite the expected increase in the computational time, the results indicate that shape optimization, which require large numbers of grid points can be resolved with a gradient-based approach.

Pandya, Mohagna J.↗
NASA NTRS1990

Continuation methods in multiobjective optimization for combined structure control design

A homotopy approach involving multiobjective functions is developed to outline the methods that have evolved for the combined control-structure optimization of physical systems encountered in the technology of large space structures. A method to effect a timely consideration of the control performance prior to the finalization of the structural design involves integrating the control and structural design processes into a unified design methodology that combines the two optimization problems into a single formulation. This study uses the combined optimization problem as a family of weighted structural and control costs. Connections with vector optimizations are described; an analysis of the zero-set of required conditions is made, and a numerical example is given.

Milman, M.↗