SEARCH · Search NASA

Results for “parameter inference”

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.

SourceResearch connectionFrom yearThrough year

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

At least 199 records · Page 11

All sources NASA NTRS DOE OSTI

NASA NTRSDate not supplied

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

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.

Loss of Control↗

NASA NTRSDate not supplied

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

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.

Loss of Control↗

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 NTRS2010

Physics-Based Methods of Failure Analysis and Diagnostics in Human Space Flight

The Integrated Health Management (IHM) for the future aerospace systems requires to interface models of multiple subsystems in an efficient and accurate information environment at the earlier stages of system design. The complexity of modern aeronautic and aircraft systems (including e.g. the power distribution, flight control, solid and liquid motors) dictates employment of hybrid models and high-level reasoners for analysing mixed continuous and discrete information flow involving multiple modes of operation in uncertain environments, unknown state variables, heterogeneous software and hardware components. To provide the information link between key design/performance parameters and high-level reasoners we rely on development of multi-physics performance models, distributed sensors networks, and fault diagnostic and prognostic (FD&P) technologies in close collaboration with system designers. The main challenges of our research are related to the in-flight assessment of the structural stability, engine performance, and trajectory control. The main goal is to develop an intelligent IHM that not only enhances components and system reliability, but also provides a post-flight feedback helping to optimize design of the next generation of aerospace systems. Our efforts are concentrated on several directions of the research. One of the key components of our strategy is an innovative approach to the diagnostics/prognostics based on the real time dynamical inference (DI) technologies extended to encompass hybrid systems with hidden state trajectories. The major investments are into the multiphysics performance modelling that provides an access of the FD&P technologies to the main performance parameters of e.g. solid and liquid rocket motors and composite materials of the nozzle and case. Some of the recent results of our research are discussed in this chapter. We begin by introducing the problem of dynamical inference of stochastic nonlinear models and reviewing earlier results. Next, we present our analytical approach to the solution of this problem based on the path integral formulation. The resulting algorithm does not require an extensive global search for the model parameters, provides optimal compensation for the effects of dynamical noise, and is robust for a broad range of dynamical models. In the following Section the strengths of the algorithm are illustrated illustrated by inferring the parameters of the stochastic Lorenz system and comparing the results with those of earlier research. Next, we discuss a number of recent results in application to the development of the IHM for aerospace system. Firstly, we apply dynamical inference approach to a solution of classical three tank problems with mixed unknown continuous and binary parameters. The problem is considered in the context of ground support system for filling fuel tanks of liquid rocket motors. It is shown that the DI algorithm is well suited for successful solution of a hybrid version of this benchmark problem even in the presence of additional periodic and stochastic perturbation of unknown strength. Secondly, we illustrate our approach by its application to an analysis of the nozzle fault in a solid rocket motor (SRM). The internal ballistics of the SRM is modelled as a set of one-dimensional partial differential equations coupled to the dynamics of the propellant regression. In this example we are specifically focussed on the inference of discrete and continuous parameters of the nozzle blocking fault and on the possibility of an application of the DI algorithm to reducing the probability of "misses" of an on-board FD&P for SRM. In the next section re-contact problem caused by first stage/upper stage separation failure is discussed. The reaction forces imposed on the nozzle of the upper stage during the re-contact and their connection to the nozzle damage and to the thrust vector control (TVC) signal are obtained. It is shown that transient impact induced torquean be modelled as a response of an effective damped oscillator. A possible application of the DI algorithm to the inference of damage parameters and predicting fault dynamics ahead of time using the actuator signal is discussed. Finally, we formulate Bayesian inferential framework for development of the IHM system for in-flight structural health monitoring (SHM) of composite materials. We consider the signal generated by piezoelectric actuator mounted on composite structure generating elastic waves in it. The signal received by the sensor is than compared with the baseline signal. The possibility of damage inference is discussed in the context of development of the SHM.

Smelyanskiy, Vadim N.↗

NASA NTRS1958

Vehicle Motions as Inferred from Radio-signal- Strength Records

Considerable data on various parameters of a satellite and its launching vehicle can be obtained from the received-signal-strength records at the various satellite receiver stations. The doppler shifts of the satellite can be used to determine the orbital characteristics quite accurately. (Ref. 1 discusses the rapid determination of these parameters.) Information concerning the velocity increment and direction of each stage can be obtained if doppler data are received by several stations during the high-speed staging. The variations in signal strength during launch can be used to find the spin rate of the satellite before, during, and after staging, and to find variations in attitude, including precession, during launch. With spin and precession frequencies available, any changes in the ratios of moments of inertia can be determined. In this paper the launchings of the Explorer satellites will be used as a basis for the generalized conclusions concerning utilization of radio-signal-strength records. The discussion includes a description of the Explorer launching system.

Pilkington, W. C.↗

NASA NTRS2004

On Bayesian Inductive Inference & Predictive Estimation

We investigate Bayesian inference and the Principle of Maximum Entropy (PME) as methods for doing inference under uncertainty. This investigation is primarily through concrete examples that have been previously investigated in the literature. We find that it is possible to do Bayesian inference and PME inference using the same information, despite claims to the contrary, but that the results are not directly comparable. This is because Bayesian inference yields a probability density function (pdf) over the unknown model parameters, whereas PME yields point estimates. If mean estimates are extracted from the Bayesian pdfs, the resulting parameter estimates can differ radically from the PME values and also from the Maximum Likelihood values. We conclude that these differences are due to the Bayesian inference not assuming anything beyond the given prior probabilities and the data, whereas PME implicitly assumes that the given constraints are the only constraints that are operating. Since this assumption can be wrong, PME values may have to be revised when subsequent data shows evidence for more constraints. The entropy concentration previously "proved" by E. T. Jaynes is shown to be in error. Further, we show that PME is a generalized form of independence assumption, and so can be a very powerful method of inference when the variables being investigated are largely independent of each other.

Cheeseman, Peter↗

NASA NTRS1997

Snow Parameters Derived from Microwave Measurements During the BOREAS Winter Field Campaign

Passive microwave data have been used to infer the snow-covered area and snow water equivalent (SWE) over forested areas, but the accuracy of these retrieved snow parameters cannot be easily validated for heterogeneous vegetated regions. The Boreal Ecosystem-Atmosphere Study Winter Field Campaign provided the opportunity to study the effect of boreal forests on snow parameter retrieval in detail. Microwave radiometers (18, 37, and 92 GHz) were flown on board the Canadian National Aeronautical Establishment's Twin Otter. Flight lines covered both the southern study area near Prince Albert and the northern study area near Thompson, Canada. During the 1994 winter campaign, extensive ground-based snow cover information, including depth, density, and grain size, was collected along most of the flight lines, jointly by U.S. and Canadian investigators. Satellite data collected by the special sensor microwave imager are also used for comparison. Preliminary results reconfirmed the relationship between microwave brightness temperature and SWE. However, the effect of forest cover observed by the aircraft sensors is different from that of the satellite observations. This is probably due to the difference in footprint averaging. There were also several flight lines flown over Candle Lake and Waskesiu Lake to assess lake ice signatures. Preliminary results show the thickness of the lake ice may be inferred from the airborne microwave observations. The microwave signature relationship between lake ice and snow matches the results from radiative transfer calculations.

Chang, A. T. C.↗

NASA NTRS1974

Reflection of a thick planar shock wave from a coplanar surface

The reflection of a thick shock wave from an adsorbing surface has been studied for argon and neon test gases using the direct simulation Monte Carlo method and laboratory shock-tube experiments. In the simulation, collisions between gas atoms were calculated with a point-center repulsive intermolecular potential, and collisions between gas atoms and the surface were described by a simple two-parameter model which includes the effects of adsorption and thermal accommodation. Experimental data included normal stress and temperature time histories on the surface of a shock-tube and wall. These data and published data of density variations near the end wall were compared with simulation results to assess the influence of adsorption and accommodation on the flowfield and the surface and also to infer optimum values for the surface-interaction model parameters. Good quantitative agreement between simulation and laboratory data was found with this simple surface-interaction model.

Deiwert, G. S.↗

NASA NTRS1989

U. radio emission from quiescent filaments

Full-disk Very Large Array (VLA) synthesis maps of the quiet Sun indicate that filaments can be seen in emission at 91.6 cm wavelength; they are detected in absorption at shorter microwave wavelengths. The 91.6 cm emission has a brightness temperature of T sub B = 3 x 10(exp 5) K. It is hotter, wider and longer than the underlying filament detected at H alpha wavelengths, but the similarity between the shape, position, elongation and orientation of the radio and optical features suggests their close association. The 91.6 cm emission is attributed to the thermal-bremsstrahlung of a hot transition sheath that envelopes the H alpha filament and acts as an interface between the cool, dense H alpha filament and the hotter, rarefied corona. The transition sheath is seen in emission because of the lower optical depth of the corona at 90 cm wavelength, and the width of this sheet is 10(exp 9) cm. A power law gradient in pressure provides a better match to the observations than a constant pressure model; definitive tests of theoretical models await simultaneous multi-wavelength studies of filaments at different observing angles. When the thermal bremsstrahlung is optically thin, the magnetic field strength in the transition sheath can be inferred from the observed circular polarization. Variable physical parameters of the sheath, such as width, electron density, and electron temperature, can explain controversial reports of the detection of, or the failure to detect, the meter-wavelength counterpart of H alpha filaments.

Lang, Kenneth R.↗

NASA NTRS2002

Broadband Observations and Modeling of the Shell-Type Supernova Remnant G347.3-0.5

The supernova remnant G347.3-0.5 emits a featureless power law in X-rays, thought to indicate shock acceleration of electrons to high energies. We here produce a broadband spectrum of the bright northwest limb of this source by combining radio observations from the Australia Telescope Compact Array (ATCA), X-ray observations from the Advanced Satellite for Cosmology and Astrophysics (ASCA), and TeV gamma-ray observations from the CANGAROO imaging Cerenkov telescope. We assume that this emission is produced by an electron population generated by diffusive shock acceleration at the remnant forward shock. The nonlinear aspects of the particle acceleration force a connection between the widely different wavelength bands and between the electrons and the unseen ions, presumably accelerated simultaneously with the electrons. This allows us to infer the relativistic proton spectrum and estimate ambient parameters such as the supernova explosion energy, magnetic field, matter density in the emission region, and efficiency of the shock acceleration process. We find convincing evidence that the shock acceleration is efficient, placing greater than 25% of the shock kinetic energy flux into relativistic ions. Despite this high efficiency, the maximum electron and proton energies, while depending somewhat on assumptions for the compression of the magnetic field in the shock, are well below the observed 'knee' at 10(exp 15) eV in the Galactic cosmic-ray spectrum.

Ellison, Donald C.↗

NASA NTRS2002

Microphysics of Waves and Instabilities in the Solar Wind and Their Macro Manifestations in the Corona and Interplanetary Space

A breakthrough resulted from the investigation of plasma heating by ion-cyclotron waves in rapidly expanding flow tubes, often referred to as coronal funnels and expected to originate in the low transition region where the temperature gradient in the solar atmosphere rises very sharply. The investigation was made using a three-fluid plasma consisting of protons, electrons and alpha particles. It was found that these waves heat the solar wind plasma by directly heating the heavier species, namely the alphas. Although only alpha particles dissipate the waves, the strong Coulomb coupling between alpha particles and protons, and between protons and electrons, makes it possible for protons and electrons to be heated also to more than one million degrees, i.e. to coronal temperatures. Interestingly, the extreme heating of the alpha particles, however, is such that they end up being hotter and faster, and are no longer in thermal equilibrium with the protons and electrons. once the rapid expansion of the flow tube is complete, the particles return to thermal equilibrium. The observational signatures of these results are such that if spectral lines formed below a million degrees are observed at different heights in the corona, the inferred outflow velocities may vary by a factor of 5 to 6. In addition, if minor ions are indeed much faster than protons and electrons at temperatures below a million degrees, then one cannot reliably determine the bulk outflow velocity of the solar wind in that region from inferences of minor ion outflow velocities. A detailed parameter study on the resonant interaction between ion cyclotron waves and alpha particles and other minor ions is also under way. It is found that the highest frequency of the waves, or the location of the heating determines whether a thermal equilibrium can be established between minor ions and protons. As the maximum frequency is increased, alpha particles and protons become farther away from thermal equilibrium. As the maximum frequency becomes smaller, alpha particles and protons are more likely to be in thermal equilibrium. However, if the maximum frequency is below a critical value, ion cyclotron resonance occurs at a greater height and a hot corona cannot be produced.

Habbal, Shadia Rifai↗

NASA NTRS2003

Coherent Anti-Stokes Raman Scattering (CARS) as a Probe for Supersonic Hydrogen-Fuel/Air Mixing

The dual-pump coherent anti-Stokes Raman spectroscopy (CARS) method was used to measure temperature and the absolute mole fractions of N2, O2 and H2 in a supersonic non-reacting fuel-air mixing experiment. Experiments were conducted in NASA Langley Research Center s Direct Connect Supersonic Combustion Test Facility. Under normal operation of this facility, hydrogen and air burn to increase the enthalpy of the test gas and O2 is added to simulate air. This gas is expanded through a Mach 2 nozzle and into a combustor model where fuel is then injected, mixes and burns. In the present experiment the O2 of the test gas is replaced by N2. The lack of oxidizer inhibited combustion of the injected H2 fuel jet allowing the fuel/air mixing process to be studied. CARS measurements were performed 427 mm downstream of the nozzle exit and 260 mm downstream of the fuel injector. Maps were obtained of the mean temperature, as well as the N2, O2 and H2 mean mole fraction fields. A map of mean H2O vapor mole fraction was also inferred from these measurements. Correlations between different measured parameters and their fluctuations are presented. The CARS measurements are compared with a preliminary computational prediction of the flow.

Danehy, P. M.↗

NASA NTRS2004

Long-Lag, Wide-pulse Gamma-Ray Bursts

Currently, the best available probe of the early phase of gamma-ray burst (GRB) jet attributes is the prompt gamma-ray emission, in which several intrinsic and extrinsic variables determine GRB pulse evolution. Bright, usually complex bursts have many narrow pulses that are difficult to model due to overlap. However, the relatively simple, long spectral lag, wide-pulse bursts may have simpler physics and are easier to model. In this work we analyze the temporal and spectral behavior of wide pulses in 24 long-lag bursts, using a pulse model with two shape parameters - width and asymmetry - and the Band spectral model with three shape parameters. We find that pulses in long-lag bursts are distinguished both temporally and spectrally from those in bright bursts: the pulses in long spectral lag bursts are few in number, and approximately 100 times wider (10s of seconds), have systematically lower peaks in vF(v), harder low-energy spectra and softer high-energy spectra. We find that these five pulse descriptors are essentially uncorrelated for our long-lag sample, suggesting that at least approximately 5 parameters are needed to model burst temporal and spectral behavior. However, pulse width is strongly correlated with spectral lag; hence these two parameters may be viewed as mutual surrogates. We infer that accurate formulations for estimating GRB luminosity and total energy will depend on several gamma-ray attributes, at least for long-lag bursts. The prevalence of long-lag bursts near the BATSE trigger threshold, their predominantly low vF(v) spectral peaks, and relatively steep upper power-law spectral indices indicate that Swift will detect many such bursts.

Norris, J. P.↗

NASA NTRS2021

Identifying Planetary Transit Candidates in TESS Full-frame Image Light Curves via Convolutional Neural Networks

The Transiting Exoplanet Survey Satellite(TESS)mission measured light from stars in∼75% of the sky throughout its 2 yr primary mission, resulting in millions of TESS 30-minute-cadence light curves to analyze in the search for transiting exoplanets. To search this vast data trove for transit signals, we aim to provide an approach that both is computationally efficient and produces highly performant predictions. This approach minimizes the required human search effort. We present a convolutional neural network, which we train to identify planetary transit signals and dismiss false positives. To make a prediction for a given light curve, our network requires no prior transit parameters identified using other methods. Our network performs inference on a TESS 30-minute-cadence light curve in∼5 ms on a single GPU, enabling large-scale archival searches. We present 181 new planet candidates identified by our network, which pass subsequent human vetting designed to rule out false positives.Our neural network model is additionally provided as open-source code for public use and extension

Gregory Olmschenk↗

NASA NTRS2024

Short–Period Variables in TESS Full–Frame Image Light Curves Identified via Convolutional Neural Networks

The Transiting Exoplanet Survey Satellite (TESS) mission measured light from stars in ∼85% of the sky throughout its 2 yr primary mission, resulting in millions of TESS 30-minute-cadence light curves to analyze in the search for transiting exoplanets. To search this vast data set, we aim to provide an approach that is computationally efficient, produces accurate predictions, and minimizes the required human search effort. We present a convolutional neural network that we train to identify short-period variables. To make a prediction for a given light curve, our network requires no prior target parameters identified using other methods. Our network performs inference on a TESS 30-minute-cadence light curve in ∼5 ms on a single GPU, enabling large-scale archival searches. We present a collection of 14,156 short-period variables identified by our network. The majority of our identified variables fall into two prominent populations, one of close-orbit main-sequence binaries and another of δ Scuti stars. Our neural network model and related code are additionally provided as open-source code for public use and extension.

Convolutional neural networks↗

NASA NTRS2007

On the Nature of the Variability Power Decay towards Soft Spectral States in X-Ray Binaries. Case Study in Cyg X-1

A characteristic feature of the Fourier Power Density Spectrum (PDS) observed from black hole X-ray binaries in low/hard and intermediate spectral states is a broad band-limited noise, characterized by a constant below some frequency (a "break" frequency) and a power law above this frequency. It has been shown that the variability of this type can be produced by the inward diffusion of the local driving perturbations in a bounded configuration (accretion disk or corona). In the framework of this model, the perturbation diffusion time to is related to the phenomenological break frequency, while the PDS power-law slope above the "break" is determined by the viscosity distribution over the configuration. The perturbation diffusion scenario explains the decay of the power of X-ray variability observed in a number of compact sources (containing black hole and neutron star) during an evolution of theses sources from low/hard to high/soft states. We compare the model predictions with the subset of data from Cyg X-1 collected by the Rossi X-ray Time Explorer (RXTE). Our extensive analysis of the Cyg X-1 PDSs demonstrates that the observed integrated power P(sub x), decreases approximately as a square root of the characteristic frequency of the driving oscillations v(sub dr). The RXTE observations of Cyg X-1 allow us to infer P(sub dr), and t(sub o) as a function of v(sub dr). We also apply the basic parameters of observed PDSs, power-law index and low frequency quasiperiodic oscillations. to infer Reynolds (Re) number from the observations using the method developed in our previous paper. Our analysis shows that Re-number increases from values about 10 in low/hard state to that about 70 during the high/soft state. Subject headings: accretion, accretion disks-black hole physics-stars:individual (Cyg X-1) :radiation mechanisms: nonthermal-physical data and processes

Titarchuk, Lev↗

NASA NTRS2002

Use of Satellite Data Assimilation to Infer Land Surface Thermal Inertia

There are two important but observationally uncertain parameters in the grid averaged surface energy budgets of mesoscale models - surface moisture availability and thermal heat capacity. A technique has been successfully developed for assimilating Geostationary Operational Environmental Satellite (GOES) skin temperature tendencies during the mid-morning time frame to improve specification of surface moisture. In a new application of the technique, the use of satellite skin temperature tendencies in early evening is explored to improve specification of the surface thermal heat capacity. Together, these two satellite assimilation constraints have been shown to significantly improve the characterization of the surface energy budget of a mesoscale model on fine spatial scales. The GOES assimilation without the adjusted heat capacity was run operationally during the International H2O Project on a 12-km grid. This paper presents the results obtained when using both the moisture availability and heat capacity retrievals in concert. Preliminary results indicate that retrieved moisture availability alone improved the verification statistics of 2-meter temperature and dew point forecasts. Results from the 1.5 month long study period using the bulk heat capacity will be presented at the meeting.

Lapenta, William↗

NASA NTRS2015

Prediction of Tensile Strength of Friction Stir Weld Joints with Adaptive Neuro-Fuzzy Inference System (ANFIS) and Neural Network

Friction-stir-welding (FSW) is a solid-state joining process where joint properties are dependent on welding process parameters. In the current study three critical process parameters including spindle speed (𝑁), plunge force (𝐹𝑧), and welding speed (𝑉) are considered key factors in the determination of ultimate tensile strength (UTS) of welded aluminum alloy joints. A total of 73 weld schedules were welded and tensile properties were subsequently obtained experimentally. It is observed that all three process parameters have direct influence on UTS of the welded joints. Utilizing experimental data, an optimized adaptive neuro-fuzzy inference system (ANFIS) model has been developed to predict UTS of FSW joints. A total of 1200 models were developed by varying the number of membership functions (MFs), type of MFs, and combination of four input variables (𝑁,𝑉,𝐹𝑧,𝐸𝐹𝐼) utilizing a MATLAB platform. Note EFI denotes an empirical force index derived from the three process parameters. For comparison, optimized artificial neural network (ANN) models were also developed to predict UTS from FSW process parameters. By comparing ANFIS and ANN predicted results, it was found that optimized ANFIS models provide better results than ANN. This newly developed best ANFIS model could be utilized for prediction of UTS of FSW joints.

Mohammad W Dewan↗