Search NASA⌕ Search

SEARCH · Search NASA

Results for “randomized response”

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 379 records · Page 21
All sourcesNASA NTRSDOE OSTI
NASA NTRS1990

First-passage problems: A probabilistic dynamic analysis for degraded structures

Structures subjected to random excitations with uncertain system parameters degraded by surrounding environments (a random time history) are studied. Methods are developed to determine the statistics of dynamic responses, such as the time-varying mean, the standard deviation, the autocorrelation functions, and the joint probability density function of any response and its derivative. Moreover, the first-passage problems with deterministic and stationary/evolutionary random barriers are evaluated. The time-varying (joint) mean crossing rate and the probability density function of the first-passage time for various random barriers are derived.

Shiao, Michael C.↗
NASA NTRS1992

Stochastic sensitivity measure for mistuned high-performance turbines

A stochastic measure of sensitivity is developed in order to predict the effects of small random blade mistuning on the dynamic aeroelastic response of turbomachinery blade assemblies. This sensitivity measure is based solely on the nominal system design (i.e., on tuned system information), which makes it extremely easy and inexpensive to calculate. The measure has the potential to become a valuable design tool that will enable designers to evaluate mistuning effects at a preliminary design stage and thus assess the need for a full mistuned rotor analysis. The predictive capability of the sensitivity measure is illustrated by examining the effects of mistuning on the aeroelastic modes of the first stage of the oxidizer turbopump in the Space Shuttle Main Engine. Results from a full analysis mistuned systems confirm that the simple stochastic sensitivity measure predicts consistently the drastic changes due to misturning and the localization of aeroelastic vibration to a few blades.

Murthy, Durbha V.↗
NASA NTRS1995

Some wave-particle effects on large-scale Alfven wave propagation and damping

Phase mixing can reduce greatly the torsional Alfven wave's dissipation length for propagation in complex magnetic field-line geometries. This phase mixing causes significant energy transfers from large to small scales where a conversion from ordered wave energy into a particle kinetic form occurs. This conversion during its initial stages is an entropy conserving process well described by Vlasov theory, Nonlinear stages of wave-particle resonance, particle trapping, and collisional resistivity are often invoked as processes eventually responsible for converting ordered wave motions into random thermal motion. Strictly speaking, this entropy producing phase cannot be described within Vlasov theory, and the large-scale effects of these microscopic events resides at the difficult frontier between generalized fluid and kinetic theories. We attempt to describe certain aspects of such resonances within the framework of fluid theory focusing on torsional Alfven wave energy transport and deposition within flux tubes.

Siregar, E.↗
NASA NTRS2001

Oxytropism: a new twist in pollen tube orientation

Chemical gradients and structural features within the pistil have been previously proposed as factors determining the directionality of pollen tube growth. In this study, we examine the behavior of pollen of eight species germinated in a dynamic oxygen gradient. While the germination rates of some species decreased directly with decreasing oxygen tension, other species showed no decrease in germination at oxygen tensions as low as 2 kPa. In one species, germination was consistently greater at decreased oxygen tensions than at ambient atmospheric levels. In three of the eight species tested, the developing pollen tube showed clear directional growth away from the more-oxygenated regions of the growth medium, while in one species growth was towards the more-oxygenated region. The remaining four species showed random tube growth. The pattern of oxytropic responses among the taxa suggests that this tropic behavior is both widespread and phylogenetically unpredictable.

NASA Discipline Plant Biology↗
NASA NTRS1998

EMG responses to maintain stance during multidirectional surface translations

To characterize muscle synergy organization underlying multidirectional control of stance posture, electromyographic activity was recorded from 11 lower limb and trunk muscles of 7 healthy subjects while they were subjected to horizontal surface translations in 12 different, randomly presented directions. The latency and amplitude of muscle responses were quantified for each perturbation direction. Tuning curves for each muscle were examined to relate the amplitude of the muscle response to the direction of surface translation. The latencies of responses for the shank and thigh muscles were constant, regardless of perturbation direction. In contrast, the latencies for another thigh [tensor fascia latae (TFL)] and two trunk muscles [rectus abdominis (RAB) and erector spinae (ESP)] were either early or late, depending on the perturbation direction. These three muscles with direction-specific latencies may play different roles in postural control as prime movers or as stabilizers for different translation directions, depending on the timing of recruitment. Most muscle tuning curves were within one quadrant, having one direction of maximal activity, generally in response to diagonal surface translations. Two trunk muscles (RAB and ESP) and two lower limb muscles (semimembranosus and peroneus longus) had bipolar tuning curves, with two different directions of maximal activity, suggesting that these muscle can play different roles as part of different synergies, depending on translation direction. Muscle tuning curves tended to group into one of three regions in response to 12 different directions of perturbations. Two muscles [rectus femoris (RFM) and TFL] were maximally active in response to lateral surface translations. The remaining muscles clustered into one of two diagonal regions. The diagonal regions corresponded to the two primary directions of active horizontal force vector responses. Two muscles (RFM and adductor longus) were maximally active orthogonal to their predicted direction of maximal activity based on anatomic orientation. Some of the muscles in each of the synergic regions were not anatomic synergists, suggesting a complex central organization for recruitment of muscles. The results suggest that neither a simple reflex mechanism nor a fixed muscle synergy organization is adequate to explain the muscle activation patterns observed in this postural control task. Our results are consistent with a centrally mediated pattern of muscle latencies combined with peripheral influence on muscle magnitude. We suggest that a flexible continuum of muscle synergies that are modifiable in a task-dependent manner be used for equilibrium control in stance.

NASA Discipline Neuroscience↗
NASA NTRS2015

Experiment Design for Complex VTOL Aircraft with Distributed Propulsion and Tilt Wing

Selected experimental results from a wind tunnel study of a subscale VTOL concept with distributed propulsion and tilt lifting surfaces are presented. The vehicle complexity and automated test facility were ideal for use with a randomized designed experiment. Design of Experiments and Response Surface Methods were invoked to produce run efficient, statistically rigorous regression models with minimized prediction error. Static tests were conducted at the NASA Langley 12-Foot Low-Speed Tunnel to model all six aerodynamic coefficients over a large flight envelope. This work supports investigations at NASA Langley in developing advanced configurations, simulations, and advanced control systems.

Murphy, Patrick C.↗
NASA NTRSDate not supplied

Nanoflare heating in coronal X-ray Bright Points

Understanding the heating of the non-flaring solar corona is an active topic of research in heliophysics. It is well accepted that magnetic fields are mainly responsible for coronal heating. The photospheric driver randomly moves the foot-points of the magnetic field lines, and either generates waves or the quasi-static buildup of magnetic energy, depending on the timescale of motion. Heating by the dissipation of the magnetic energy is termed as DC heating while the dissipation of wave is known as AC heating mechanism. Both the AC and DC heating mechanisms can lead to impulsive heating events, termed nanoflares. The magnitude and frequency of these nanoflares determine whether they can adequately satisfy the coronal heating budget. Thus it is of great importance to study the nanoflares properties to validate their role in coronal heating. Individual nanoflares are difficult to detect with the present generation instruments, however their presence can be inferred through indirect observational techniques combined with simulated nanoflare-heated plasma emissions. We combined the observed emission properties of coronal X-ray bright points (XBP) in EUV and X-ray wavelengths with the simulated nanoflare heated plasma to understand nanoflare properties and their contribution in coronal heating. Our results indicate that the heating of coronal XBPs could be explained by frequent nanoflares. Further we demonstrate that the sensitive spectroscopic observations in X-ray wavelength are crucial to diagnose the nanoflare properties.

Biswajit Mondal↗
NASA NTRS2005

A First and Second Order Moment Approach to Probabilistic Control Synthesis

This paper presents a robust control design methodology based on the estimation of the first two order moments of the random variables and processes that describe the controlled response. Synthesis is performed by solving an multi-objective optimization problem where stability and performance requirements in time- and frequency domains are integrated. The use of the first two order moments allows for the efficient estimation of the cost function thus for a faster synthesis algorithm. While reliability requirements are taken into account by using bounds to failure probabilities, requirements related to undesirable variability are implemented by quantifying the concentration of the random outcome about a deterministic target. The Hammersley Sequence Sampling and the First- and Second-Moment- Second-Order approximations are used to estimate the moments, whose accuracy and associated computational complexity are compared numerically. Examples using output-feedback and full-state feedback with state estimation are used to demonstrate the ideas proposed.

Crespo, Luis G.↗
DOE OSTI2019

Total experimental uncertainty in hydrodynamic testing of a semisubmersible wind turbine, considering numerical propagation of systematic uncertainty

Quantifying the uncertainty in experimental results is a vital step in properly validating numerical simulation tools for designing floating wind turbines; without a good understanding of the experimental uncertainties, it is impossible to confirm if numerical simulation tools can capture the physics with acceptable accuracy. Recent validation studies suggest that the wave-induced, low-frequency surge and pitch motions of semisubmersible-type floating wind turbines are consistently underpredicted by numerical simulations, but it has not been possible to state whether or not this underprediction is within the level of experimental error. Here, previously assessed systematic uncertainty components in hydrodynamic tests of the OC5-DeepCwind semisubmersible are propagated to response metrics of interest using numerical simulation tools, and combined with the system's random uncertainty to obtain the total experimental uncertainty. The uncertainty in the low-frequency response metrics is found to be most sensitive to the system properties (e.g., mooring stiffness and center of gravity), and also the wave elevation. The results of the present study suggest that the underprediction of the low-frequency response behavior observed in previous validation studies is larger than the experimental uncertainty.

17 WIND ENERGY↗
DOE OSTI2024

Polynomial chaos expansions on principal geodesic Grassmannian submanifolds for surrogate modeling and uncertainty quantification

In this work we introduce a manifold learning-based surrogate modeling framework for uncertainty quantification in high-dimensional stochastic systems. Our first goal is to perform data mining on the available simulation data to identify a set of low-dimensional (latent) descriptors that efficiently parameterize the response of the high-dimensional computational model. To this end, we employ Principal Geodesic Analysis on the Grassmann manifold of the response to identify a set of disjoint principal geodesic submanifolds, of possibly different dimension, that captures the variation in the data. Since operations on the Grassmann require the data to be concentrated, we propose an adaptive algorithm based on Riemannian K-means and the minimization of the sample Fréchet variance on the Grassmann manifold to identify “local” principal geodesic submanifolds that represent different system behavior across the parameter space. Polynomial chaos expansion is then used to construct a mapping between the random input parameters and the projection of the response on these local principal geodesic submanifolds. Here, the method is demonstrated on four test cases, a toy-example that involves points on a hypersphere, a Lotka-Volterra dynamical system, a continuous-flow stirred-tank chemical reactor system, and a two-dimensional Rayleigh-Bénard convection problem.

42 ENGINEERING↗
NASA NTRS2009

Toward Online Measurement of Decision State

In traditional perceptual decision-making experiments, two pieces of data are collected on each trial: response time and accuracy. But how confident were participants and how did their decision state evolve over time? We asked participants to provide a continuous readout of their decision state by moving a cursor along a sliding scale between a 100% certain left response and a 100% certain right response. Subjects did not terminate the trials; rather, trials were timed out at random and subjects were scored based on the cursor position at that time. Higher rewards for correct responses and higher penalties for errors were associated with extreme responses so that the response with the highest expected value was that which accurately reflected the participant's odds of being correct. This procedure encourages participants to expose the time-course of their evolving decision state. Evidence on how well they can do this will be presented.

Lachter, Joel↗
NASA NTRS2009

Toward Online Measurement of Decision State

In traditional perceptual decision-making experiments, two pieces of data recollected on each trial: response time and accuracy. But how confident were participants and how did their decision state evolve over time? We asked participants to provide a continuous readout of their decision state by moving a cursor along a sliding scale between a 100% certain left response and a 100% certain right response. Subjects did not terminate the trials; rather, trials were timed out at random and subjects were scored based on the cursor position at the time. Higher rewards for correct responses and higher penalties for errors were associated with extreme responses so that the response with the highest ex[pected value was that which accurately reflected the participant's odds of being correct. This procedure encourages participants to expose the time-course of their evolving decision state. Evidence on how well they can do this will be presented.

Lachter, Joel↗
DOE OSTI2020

A machine learning-based fast frequency response control for a VSC-HVDC system

An HVDC system can realize a very fast frequency response to the disturbed system under a contingency because its active power control is decoupled from the frequency deviation. However, most of existing HVDC frequency control strategies are coupled with system primary frequency control and secondary frequency control. Since the traditional system frequency control is dominated by the thermal generators, the advantage of the fast response of the HVDC system is not made fully used. The development of a frequency response estimation based on a machine learning algorithm provides another approach to improve the frequency response capability of the HVDC system. Different from other frequency deviation tracking strategies, a machine learning based HVDC frequency response control can directly increase the power flow of a HVDC system by estimation of the system generator or load lost. In this paper, a fast frequency response control using a HVDC system for a large power system disturbance based on the multivariate random forest regression (MRFR) algorithm is proposed. The simulation is carried out with an integrated power system model based on the North American interconnections. The simulation results indicate that the proposed MRFR based frequency response control can significantly improve the frequency low point during an event, while stabilizing the frequency in advance.

42 ENGINEERING↗
NASA NTRS1973

Nonstationary ascent wind analysis.

A statistical approach for computing launch vehicle response to inflight winds is presented. In the response analysis the zonal component of the ascent wind velocity is modeled as a nonstationary random process. Model parameters are computed by processing available measured wind data. Pertinent pitch plane response statistics are computed and their value in constructing fatigue load spectra for a reusable launch vehicle is demonstrated. A method for using the response statistics to estimate structural reliability is also discussed. The accuracy and the cost saving potential of the probabilistic structural dynamics approach are assessed by performing comparisons with digital simulation results.

Howell, L. J.↗
NASA NTRS1990

Active control of multi-dimensional random sound in ducts

Previous work has demonstrated how active control may be applied to the control of random noise in ducts. These implementations, however, have been restricted to frequencies where only plane waves are propagating in the duct. In spite of this, the need for this technology at low frequencies has progressed to the point where commercial products that apply these concepts are currently available. Extending the frequency range of this technology requires the extension of current single channel controllers to multi-variate control systems as well as addressing the problems inherent in controlling higher order modes. The application of active control in the multi-dimensional propagation of random noise in waveguides is examined. An adaptive system is implemented using measured system frequency response functions. Experimental results are presented illustrating attained suppressions of 15 to 30 dB for random noise propagating in multiple modes.

Silcox, R. J.↗
DOE OSTI2021

Design Space Exploration of Emerging Memory Technologies for Machine Learning Applications

Memory design space exploration methods study memory systems’ performances and limitations before implementation. The computer memory design space has grown exponentially because of the enormous growth of memory types, memory controllers, and application software. Computer simulators are commonly used for memory design space exploration. However, complex memory simulations take an enormous amount of time. Hence, in this paper, we proposed a machine learning-based design space exploration method for dynamic random-access memory and non-volatile memory systems. We applied our method to the CosmoGAN and LeNet applications to predict the following six memory response parameters: (i) bandwidth, (ii) power, (iii) average latency, (iv) average total latency, (v) memory reads, and (vi) memory writes. Our experimental results show that machine learning models can predict memory response parameter values faster than simulations. We used support vector machine, random forest, and gradient boosting machine learning models. We observed that the support vector machine provides better performance for bandwidth, average latency, and average total latency. The random forest model works better for memory reads and writes. The gradient boosting model provides superior prediction performance for power. We provide a detailed discussion on learning curve characteristics, error analysis, and memory type recommendation.

Hasan, S M Shamimul↗
NASA NTRS2002

Selective norepinephrine reuptake inhibition as a human model of orthostatic intolerance

BACKGROUND: Observations in patients with functional mutations of the norepinephrine transporter (NET) gene suggest that impaired norepinephrine uptake may contribute to idiopathic orthostatic intolerance. METHODS AND RESULTS: We studied the effect of the selective NET blocker reboxetine and placebo in a randomized, double-blind, crossover fashion on cardiovascular responses to cold pressor testing, handgrip testing, and a graded head-up tilt test (HUT) in 18 healthy subjects. In a subset, we determined isoproterenol and phenylephrine sensitivities. Subjects ingested 8 mg reboxetine or placebo 12 hours and 1 hour before testing. In the supine position, heart rate was 65+/-2 bpm with placebo and 71+/-3 bpm with reboxetine. At 75 degrees HUT, heart rate was 84+/-3 and 119+/-4 bpm with placebo and with reboxetine (P<0.0001). Mean arterial pressure was 85+/-2 with placebo and 91+/-2 mm Hg with reboxetine while supine (P<0.01) and 88+/-2 mm Hg and 90+/-3 mm Hg at 75 degrees HUT. Blood pressure responses to cold pressor and handgrip testing were attenuated with reboxetine. Reboxetine increased the sensitivity to the chronotropic effect of isoproterenol and the pressor effect of phenylephrine. Vasovagal reactions occurred in 9 subjects on placebo and in 1 subject on reboxetine. CONCLUSIONS: Selective NET blockade creates a phenotype that resembles idiopathic orthostatic intolerance. This observation supports the hypothesis that disordered norepinephrine uptake mechanisms can contribute to human cardiovascular disease. Our study also suggests that NET inhibition might be useful in preventing vasovagal reactions.

Non-NASA Center↗