Search NASA⌕ Search

SEARCH · Search NASA

Results for “variational optimization”

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 721 records · Page 40

Primal and dual formulations of sequential gradient-restoration algorithms for trajectory optimization problems

One of the most effective first-order algorithms for solving trajectory optimization problems is the sequential gradient-restoration algorithm (SGRA). Originally developed in the primal formulation, this algorithm is extended to incorporate a dual formulation. Both the primal formulation and the dual formulation involve a sequence of two-phase cycles, each cycle including a gradient phase and a restoration phase. In turn, each iteration of the gradient phase and the restoration phase requires the solution of an auxiliary minimization problem (AMP). In the primal formulation, the AMP is solved with respect to the variations of the state, the control, and the parameter. In the dual formulation, the AMP is solved with respect to the Lagrange multipliers. A characteristic of the dual formulation is that the AMPs associated with the gradient phase and the restoration phase of SGRA can be reduced to mathematical programming problems involving a finite number of parameters as unknowns. A comparison of the primal formulation and the dual formulation is presented. The comparison is done in terms of several trajectory optimization problems having current aerospace interest.

Miele, A.↗

Complete exchange on the iPSC-860

The implementation of complete exchange on the circuit switched Intel iPSC-860 hypercube is described. This pattern, also known as all-to-all personalized communication, is the densest requirement that can be imposed on a network. On the iPSC-860, care needs to be taken to avoid edge contention, which can have a disastrous impact on communication time. There are basically two classes of algorithms that achieve contention-free complete exchange. The first contains the classical standard exchange algorithm that is generally useful for small message sizes. The second includes a number of optimal or near-optimal algorithms that are best for large messages. Measurement of communication overhead on the iPSC-860 are given and a notation for analyzing communication link usage is developed. It is shown that for the two classes of algorithms, there is substantial variation in performance with synchronization technique and choice of message protocol. Timings of six implementations are given; each of these is useful over a particular range of message size and cube dimension. Since the complete exchange is a superset of communication patterns, these timings represent upper bounds on the time required by an arbitrary communication requirement. These results indicate that the programmer needs to evaluate several possibilities before finalizing an implementation - a careful choice can lead to very significant savings in time.

Bokhari, Shahid H.↗

Parametric Geometry, Structured Grid Generation, and Initial Design Study for REST-Class Hypersonic Inlets

Computational design and analysis of three-dimensional hypersonic inlets with shape transition has been a significant challenge due to the complex geometry and grid required for three-dimensional viscous flow calculations. Currently, the design process utilizes an inviscid design tool to produce initial inlet shapes by streamline tracing through an axisymmetric compression field. However, the shape is defined by a large number of points rather than a continuous surface and lacks important features such as blunt leading edges. Therefore, a design system has been developed to parametrically construct true CAD geometry and link the topology of a structured grid to the geometry. The Adaptive Modeling Language (AML) constitutes the underlying framework that is used to build the geometry and grid topology. Parameterization of the CAD geometry allows the inlet shapes produced by the inviscid design tool to be generated, but also allows a great deal of flexibility to modify the shape to account for three-dimensional viscous effects. By linking the grid topology to the parametric geometry, the GridPro grid generation software can be used efficiently to produce a smooth hexahedral multiblock grid. To demonstrate the new capability, a matrix of inlets were designed by varying four geometry parameters in the inviscid design tool. The goals of the initial design study were to explore inviscid design tool geometry variations with a three-dimensional analysis approach, demonstrate a solution rate which would enable the use of high-fidelity viscous three-dimensional CFD in future design efforts, process the results for important performance parameters, and perform a sample optimization.

Ferlemann, Paul G.↗

Stationkeeping Monte Carlo Simulation for the James Webb Space Telescope

The James Webb Space Telescope (JWST) is scheduled to launch in 2018 into a Libration Point Orbit (LPO) around the Sun-Earth/Moon (SEM) L2 point, with a planned mission lifetime of 10.5 years after a six-month transfer to the mission orbit. This paper discusses our approach to Stationkeeping (SK) maneuver planning to determine an adequate SK delta-V budget. The SK maneuver planning for JWST is made challenging by two factors: JWST has a large Sunshield, and JWST will be repointed regularly producing significant changes in Solar Radiation Pressure (SRP). To accurately model SRP we employ the Solar Pressure and Drag (SPAD) tool, which uses ray tracing to accurately compute SRP force as a function of attitude. As an additional challenge, the future JWST observation schedule will not be known at the time of SK maneuver planning. Thus there will be significant variation in SRP between SK maneuvers, and the future variation in SRP is unknown. We have enhanced an earlier SK simulation to create a Monte Carlo simulation that incorporates random draws for uncertainties that affect the budget, including random draws of the observation schedule. Each SK maneuver is planned to optimize delta-V magnitude, subject to constraints on spacecraft pointing. We report the results of the Monte Carlo simulations and discuss possible improvements during flight operations to reduce the SK delta-V budget.

Stationkeepting↗

James Webb Space Telescope (JWST) Stationkeeping Monte Carlo Simulations

The James Webb Space Telescope (JWST) will launch in 2018 into a Libration Point Orbit (LPO) around the Sun-EarthMoon (SEM) L2 point, with a planned mission lifetime of 11 years. This paper discusses our approach to Stationkeeping (SK) maneuver planning to determine an adequate SK delta-V budget. The SK maneuver planning for JWST is made challenging by two factors: JWST has a large Sunshield, and JWST will be repointed regularly producing significant changes in Solar Radiation Pressure (SRP). To accurately model SRP we employ the Solar Pressure and Drag (SPAD) tool, which uses ray tracing to accurately compute SRP force as a function of attitude. As an additional challenge, the future JWST observation schedule will not be known at the time of SK maneuver planning. Thus there will be significant variation in SRP between SK maneuvers, and the future variation in SRP is unknown. We have enhanced an earlier SK simulation to create a Monte Carlo simulation that incorporates random draws for uncertainties that affect the budget, including random draws of the observation schedule. Each SK maneuver is planned to optimize delta-V magnitude, subject to constraints on spacecraft pointing. We report the results of the Monte Carlo simulations and discuss possible improvements during flight operations to reduce the SK delta-V budget.

Monte Carlo↗

Neural-Based Compression Scheme for Solar Image Data

Studying the solar system and especially the Sun relies on the data gathered daily from space missions. These missions are data-intensive and compressing this data to make them efficiently transferable to the ground station is a twofold decision to make. Stronger compression methods, by distorting the data, can increase data throughput at the cost of accuracy which could affect scientific analysis of the data. On the other hand, preserving subtle details in the compressed data requires a high amount of data to be transferred, reducing the desired gains from compression. In this work, we propose a neural network-based lossy compression method to be used in NASA’s data-intensive imagery missions. We chose NASA’s Solar Dynamics Observatory (SDO) mission which transmits 1.4 terabytes of data each day as a proof of concept for the proposed algorithm. In this work, we propose an adversarially trained neural network, equipped with local and non-local attention modules to capture both the local and global structure of the image resulting in a better trade-off in rate-distortion (RD) compared to conventional hand-engineered codecs. The RD variational autoencoder used in this work is jointly trained with a channel-dependent entropy model as a shared prior between the analysis and synthesis transforms to make the entropy coding of the latent code more effective. We also studied how optimizing perceptual losses could help our neural compressor to preserve high-frequency details of the data in the reconstructed compressed image. Our neural image compression algorithm outperforms currently-in-use and state-of-the-art codecs such as JPEG and JPEG-2000 in terms of the RD performance when compressing extreme-ultraviolet (EUV) data. As a proof of concept for use of this algorithm in SDO data analysis, we have performed coronal hole (CH) detection using our compressed images, and generated consistent segmentations, even at a compression rate of ∼ 0.1 bits per pixel (compared to 8 bits per pixel on the original data) using EUV data from SDO.

Image coding↗

Design of a Variational Multiscale Method for Turbulent Compressible Flows

A spectral-element framework is presented for the simulation of subsonic compressible high-Reynolds-number flows. The focus of the work is maximizing the efficiency of the computational schemes to enable unsteady simulations with a large number of spatial and temporal degrees of freedom. A collocation scheme is combined with optimized computational kernels to provide a residual evaluation with computational cost independent of order of accuracy up to 16th order. The optimized residual routines are used to develop a low-memory implicit scheme based on a matrix-free Newton-Krylov method. A preconditioner based on the finite-difference diagonalized ADI scheme is developed which maintains the low memory of the matrix-free implicit solver, while providing improved convergence properties. Emphasis on low memory usage throughout the solver development is leveraged to implement a coupled space-time DG solver which may offer further efficiency gains through adaptivity in both space and time.

Design↗

A Study of the Impact of Variations on Aerodynamic Flow in Gas Turbine Engines via Monte-Carlo Simulations

The unsteady compressible inviscid flow is characterized by the conservations of mass, momentum, and energy; or simply the Euler equations. In this paper, a study of the subsonic one-dimensional Euler equations with local preconditioning is presented with a modal analysis approach. Specifically, this study investigates the behavior of airflow in a gas turbine engine using the specified conditions at the inflow and outflow boundaries of the compressor, combustion chamber, and turbine, under the impact of variations in pressure, velocity, temperature, and density at low Mach numbers. Two main questions that motivate this research are: 1) Is there any aerodynamic problem with the existing gas turbine engines that could impact aircraft performance? 2) If yes, what aspect of a gas turbine engine could be improved via design to alleviate that impact and to optimize aircraft performance. This paper presents an initial attempt to the flow behavior in terms (perturbation) using simulation outputs from a customer-deck model obtained from Pratt&Whitney, (i.e., pressure, temperature, velocity, density) about their mean states at the inflow and outflow boundaries of the compressor, combustion chamber, and turbine. Flow behavior is analyzed for the high pressure compressor and combustion chamber employing the conditions on their left and right boundaries. In the same fashion, similar analyses are carried out for the high and low-pressure turbines. In each case, the eigenfrequencies that are obtained for different boundary conditions are examined closely based on their probabilistic distributions, a result of a Monte Carlo 10,000-sample simulation. Furthermore, the characteristic waves and eave response are analyzed and contrasted among different cases, with and without preconditioners. The results reveal the existence of flow instabilities due to the combined effect of variations and excessive pressures; which are clearly the case in the combustion chamber and high-pressure turbine. Finally a discussion is presented on potential impacts of the instabilities and what can be improved via design to alleviate them for a better aircraft performance.

Ngo, Khiem Viet↗

Fine guidance sensor design optimization for Space Infrared Telescope Facility (SIRTF)

The present paper is concerned with the general issues of design optimization of a staring mosaic sensor for guide-star tracking, taking into account the choice of the proper interpolation algorithm. It was found that the final choice of algorithm depends on the severity of the positional accuracy requirement and the expected variation in the blur spot shape. The considered track algorithms estimate the position either of the peak or of the centroid of the image of a point source. Attention is given to aspects of problem definition, the total effect of error sources on algorithm accuracy, a parametric analysis of performance behavior, Monte Carlo results, and the design of the SIRTF Fine Guidance Sensor (FGS). The algorithm error, the signal-to-noise error, and the nonuniformity error are examined.

Grossman, S. B.↗

Integrated aeroservoelastic tailoring of lifting surfaces

An approach to the integration of two design activities, structural design and active control design, for a highly idealized aeroservoelastic system is presented. The particular design goal for this study is the maximization of the stable airspeed envelope of an idealized model of an aeroservoelastic system through rational and systematic variation of structural and control design parameters. The steady-state linear quadratic regulator is used to model the control subsystem; the structural subsystem is assigned characteristic design parameters, such as shear center position. The application of the procedure described here produces optimally controlled structures with stability characteristics superior to those of open-loop and initial closed-loop designs.

Zeiler, T. A.↗

Shape optimization of 3D continuum structures via force approximation techniques

The existing need to develop methods whereby the shape design efficiency can be improved through the use of high quality approximation methods is addressed. An efficient approximation method for stress constraints in 3D shape design problems is proposed based on expanding the nodal forces in Taylor series with respect to shape variations. The significance of this new method is shown through elementary beam theory calculations and via numerical computations using 3D solid finite elements. Numerical examples including the classical cantilever beam structure and realistic automotive parts like the engine connecting rod are designed for optimum shape using the proposed method. The numerical results obtained from these methods are compared with other published results, to assess the efficiency and the convergence rate of the proposed method.

Vanderplaats, Garret N.↗

Navier-Stokes computations of a viscous optimized waverider

The performance of a Mach 6 viscous optimized waverider was calculated using the 3-D Navier-Stokes equations. The Mach 6 viscous optimized waverider was generated using MAXWARP, a code developed at the University of Maryland. The computations were performed using CFL3D, an implicit upwind-biased finite-volume algorithm developed at NASA Langley. Results show that good agreement was found between the calculated performance by MAXWARP and results from the Mach 6 Navier-Stokes computation. Furthermore, off-design performance of the Mach 6 optimized waverider was computed at Mach 4 and 8. The performance at these Mach numbers compared well with the performance of the viscous optimized waveriders specifically designed for these Mach numbers. Finally, contours of different flow parameters in the cross-flow plane were examined for the three calculations. The results indicate that the flow gradients are relatively small within the captured flow, and the variation itself is well behaved; thus, making the waverider configuration a promising choice for an engine/airframe design, especially for cruise-type applications.

Takashima, Naruhisa↗

Effect of Crystal Orientation on Fatigue Failure of Single Crystal Nickel Base Turbine Blade Superalloys

High Cycle Fatigue (HCF) induced failures in aircraft gas-turbine engines is a pervasive problem affecting a wide range of components and materials. HCF is currently the primary cause of component failures in gas turbine aircraft engines. Turbine blades in high performance aircraft and rocket engines are increasingly being made of single crystal nickel superalloys. Single-crystal Nickel-base superalloys were developed to provide superior creep, stress rupture, melt resistance and thermomechanical fatigue capabilities over polycrystalline alloys previously used in the production of turbine blades and vanes. Currently the most widely used single crystal turbine blade superalloys are PWA 1480/1493 and PWA 1484. These alloys play an important role in commercial, military and space propulsion systems. PWA1493, identical to PWA1480, but with tighter chemical constituent control, is used in the NASA SSME (Space Shuttle Main Engine) alternate turbopump, a liquid hydrogen fueled rocket engine. Objectives for this paper are motivated by the need for developing failure criteria and fatigue life evaluation procedures for high temperature single crystal components, using available fatigue data and finite element modeling of turbine blades. Using the FE (finite element) stress analysis results and the fatigue life relations developed, the effect of variation of primary and secondary crystal orientations on life is determined, at critical blade locations. The most advantageous crystal orientation for a given blade design is determined. Results presented demonstrates that control of secondary and primary crystallographic orientation has the potential to optimize blade design by increasing its resistance to fatigue crack growth without adding additional weight or cost.

Arakere, Nagaraj K.↗

Helicopter Flight Procedures for Community Noise Reduction

A computationally efficient, semiempirical noise model suitable for maneuvering flight noise prediction is used to evaluate the community noise impact of practical variations on several helicopter flight procedures typical of normal operations. Turns, "quick-stops," approaches, climbs, and combinations of these maneuvers are assessed. Relatively small variations in flight procedures are shown to cause significant changes to Sound Exposure Levels over a wide area. Guidelines are developed for helicopter pilots intended to provide effective strategies for reducing the negative effects of helicopter noise on the community. Finally, direct optimization of flight trajectories is conducted to identify low noise optimal flight procedures and quantify the magnitude of community noise reductions that can be obtained through tailored helicopter flight procedures. Physically realizable optimal turns and approaches are identified that achieve global noise reductions of as much as 10 dBA Sound Exposure Level.

Greenwood, Eric↗

Constraining Numerical Geodynamo Modeling with Surface Observations

Numerical dynamo solutions have traditionally been generated entirely by a set of self-consistent differential equations that govern the spatial-temporal variation of the magnetic field, velocity field and other fields related to dynamo processes. In particular, those solutions are obtained with parameters very different from those appropriate for the Earth s core. Geophysical application of the numerical results therefore depends on correct understanding of the differences (errors) between the model outputs and the true states (truth) in the outer core. Part of the truth can be observed at the surface in the form of poloidal magnetic field. To understand these differences, or errors, we generate new initial model state (analysis) by assimilating sequentially the model outputs with the surface geomagnetic observations using an optimal interpolation scheme. The time evolution of the core state is then controlled by our MoSST core dynamics model. The final outputs (forecasts) are then compared with the surface observations as a means to test the success of the assimilation. We use the surface geomagnetic data back to year 1900 for our studies, with 5-year forecast and 20-year analysis periods. We intend to use the result; to understand time variation of the errors with the assimilation sequences, and the impact of the assimilation on other unobservable quantities, such as the toroidal field and the fluid velocity in the core.

Kuang, Weijia↗

Altitude Variation of the CO2(V2)-O Quenching Rate Coefficient in Mesosphere and Lower Thermosphere

Among the processes governing the energy balance in the mesosphere and lower thermosphere (MLT), the quenching of CO2(v2) vibrational levels by collisions with oxygen atoms plays an important role. However, the k(CO2-O) values measured in the lab and retrieved from atmospheric measurements vary from 1.5 x 10(exp -12)cu cm/s through 9.0 x 10(exp -12)cu cm/s that requires further studying. In this work we used synergistic data from a ground based lidar and a satellite infrared radiometer to estimate k(CO2-O). We used the night- and daytime temperatures between 80 and 110 km measured by the Colorado State University narrow-band sodium (Na) lidar located at Fort Collins, Colorado (41 N, 255E) as ground truth of the SABER/TIMED nearly simultaneous ( +/-10 minutes) and common volume (within +/-1 degree in latitude, +/-2 degrees in longitude) observations. For each altitude in 80-110 km interval we estimate an 'optimal" value of k(CO2-O) needed to minimize the discrepancy between the simulated 15 micron CO2 radiance and that measured by the SABER/TIMED instrument. The k(CO2-O) obtained in this way varies in altitude from 3.5 x 10(exp -12)cu cm/s at 80 km to 5.2 x 10(exp -12)cu cm/s for altitudes above 95 km. We discuss this variation of the rate constant and its impact on temperature retrievals from 15 pm radiance measurements and on the energy budget of MLT.

Feofilov, A.↗

A Look at the Truths and Misconceptions of the Variational Quantum Eigensolver and the Implications of Overparameterization

In this work, we investigate loss landscapes of the variational quantum eigensolver (VQE) by quantifying the number of local minima through empirical analyses. We focus on minimal models in chemistry and physics so that we can do a complete analysis using more computationally expensive tools. We employ Hessian eigenvalue calculations and the nudged elastic band algorithm to characterize these landscapes. Our results expand upon the existing literature by highlighting the optimization challenges faced by VQE. We find that, as the number of parameters in our ansatz increases, the number of basins increases while the corresponding loss function values converge toward the global minimum value. This observation implies that overparameterization may lead to an ``effective convexity'' in VQE loss landscapes, a phenomenon supported by theoretical and numerical work in classical machine learning.

quantum computing↗