Search NASA⌕ Search

SEARCH · Search NASA

Results for “Numerical 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 1,459 records · Page 81

Optimization-Based Azeotropic Distillation System Synthesis Using Geometric Insights

The synthesis of azeotropic distillation systems is challenging due to the existence of compartments in the residue curve map coupled with the combinatorial aspects from numerous possible system configurations. In this work, an optimization-based approach is introduced to synthesize homogeneous azeotropic distillation systems. The approach employs a network-based representation generated via a matrix method. To design the distillation columns, the modified Underwood equations are adopted, in which pseudocomponent-based compositions are used. For cases where separatrices are significantly curved, corrections via piece-wise linear functions and collinearity properties are implemented. Two examples are presented to illustrate the proposed approach.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Constrained or unconstrained? Neural-network-based equation discovery from data

Throughout many fields, practitioners often rely on differential equations to model systems. Yet, for many applications, the theoretical derivation of such equations and/or the accurate resolution of their solutions may be intractable. Instead, recently developed methods, including those based on parameter estimation, operator subset selection, and neural networks, allow for the data-driven discovery of both ordinary and partial differential equations (PDEs), on a spectrum of interpretability. The success of these strategies is often contingent upon the correct identification of representative equations from noisy observations of state variables and, as importantly and intertwined with that, the mathematical strategies utilized to enforce those equations. Specifically, the latter has been commonly addressed via unconstrained optimization strategies. Representing the PDE as a neural network, we propose to discover the PDE (or the associated operator) by solving a constrained optimization problem and using an intermediate state representation similar to a physics-informed neural network (PINN). The objective function of this constrained optimization problem promotes matching the data, while the constraints require that the discovered PDE is satisfied at a number of spatial collocation points. We present a penalty method and a widely used trust-region barrier method to solve this constrained optimization problem, and we compare these methods on numerical examples. Our results on several example problems demonstrate that the latter constrained method outperforms the penalty method, particularly for higher noise levels or fewer collocation points. This work motivates further exploration into using sophisticated constrained optimization methods in scientific machine learning, as opposed to their commonly used, penalty-method or unconstrained counterparts. For both of these methods, we solve these discovered neural network PDEs with classical methods, such as finite difference methods, as opposed to PINNs-type methods relying on automatic differentiation. Here, we briefly highlight how simultaneously fitting the data while discovering the PDE improves the robustness to noise and other small, yet crucial, implementation details.

Data-driven discovery↗

Optimizing Parameters for Deep-Space Optical Communication

A paper discusses the optimization of the parameters of a high-rate, deep-space optical communication link that utilizes pulse-position modulation (PPM) and an error-correcting code (ECC). The parameters in question include the PPM order (number of pulse time slots in one symbol period), the ECC rate, and the uncoded symbol error rate. In simple terms, the optimization problem is to choose the combination of these parameters that maximizes the throughput data rate at a given bit-error-rate (BER), subject to several constraints, including limits on the average and peak power and possibly a limit on the uncoded symbol error rate. This is a complex, multidimensional optimization problem, the solution of which involves computation of channel capacities for various combinations of the parameters. The paper presents extensive theoretical analyses and numerical predictions that elucidate the many facets of the optimization problem. It shows how a nearly optimum solution can be obtained by choosing the optimum PPM order for the desired number of bits per slot and concatenating the PPM mapping with an error-correction code so that the decoded bits satisfy some BER threshold.

Moison, Bruce↗

Multi-objective Bayesian optimization of ferroelectric materials with interfacial control for memory and energy storage applications

Optimization of materials’ performance for specific applications often requires balancing multiple aspects of materials’ functionality. Even for the cases where a generative physical model of material behavior is known and reliable, this often requires search over multidimensional function space to identify low-dimensional manifold corresponding to the required Pareto front. In this work, we introduce the multi-objective Bayesian optimization (MOBO) workflow for the ferroelectric/antiferroelectric performance optimization for memory and energy storage applications based on the numerical solution of the Ginzburg–Landau equation with electrochemical or semiconducting boundary conditions. MOBO is a low computational cost optimization tool for expensive multi-objective functions, where we update posterior surrogate Gaussian process models from prior evaluations and then select future evaluations from maximizing an acquisition function. Using the parameters for a prototype bulk antiferroelectric (PbZrO 3 ), we first develop a physics-driven decision tree of target functions from the loop structures. We further develop a physics-driven MOBO architecture to explore multidimensional parameter space and build Pareto-frontiers by maximizing two target functions jointly—energy storage and loss. This approach allows for rapid initial materials and device parameter selection for a given application and can be further expanded toward the active experiment setting. The associated notebooks provide both the tutorial on MOBO and allow us to reproduce the reported analyses and apply them to other systems (https://github.com/arpanbiswas52/MOBO_AFI_Supplements).

36 MATERIALS SCIENCE↗

Performing Numerical Analysis of Cybersecurity Options Using Dynamic Risk Analysis Tool EMRALD

Cyberattacks can have many different attack paths, durations, and goals. There are also many different mitigation options involving hardware, software, and/or humans. Considering a cyber threat should involve defense-in-depth methods and a quantitative or numerical evaluation of overall effectiveness against dynamic, time-dependent attacks to make cost and risk-informed decisions. Typical cyberattack modeling methods only provide a qualitative evaluation. The main areas of cybersecurity are confidentiality, integrity, and availability. For companies with cyber-physical systems such as advanced nuclear reactors, cyber-related safety is a requirement set by North American Electric Reliability and the U.S. Nuclear Regulatory Commission. They are also concerned about availability or reliability as a business case. As cyber threats are evolving to a business-for-hire structure, more attacks may focus on disrupting business success and reliability, causing financial and economic stability risk. Companies want to know business reliability and recovery from those threats, and that requires modeling physical behavior of the targets. Dynamic-state-based and Markov-based modeling provides a method for better cyber scenario modeling with different tools having issues such as state-base explosion. Dynamic modeling enables time and conditional features not found in other numerical evaluation methods. EMRALD (Event Modeling Risk Assessment using Lined Diagrams) is a dynamic risk analysis modeling and simulation tool and has features that reduce modeling issues. It has been used to model different time-dependent events including plant behavior and operator procedures. As a general modeling tool, EMRALD can also be used to model cyberattack scenarios with varying mitigation options and quantify effectiveness, producing numerical data for risk-informed decisions. This paper uses EMRALD to demonstrate that dynamic numerical risk analysis can be used for cyber threat modeling to provide insights for design decision-making and optimize defense strategies. Keywords: cyber modeling; cyber-physical systems; numerical cyber modeling

97 - MATHEMATICS AND COMPUTING↗

An optimal GPS data processing technique for precise positioning

A mathematical formula to optimally combine dual-frequency GPS pseudorange and carrier phase (integrated Doppler) data streams into a single data stream is derived in closed form. The data combination reduces the data volume and computing time in the filtering process for parameter estimation by a factor of 4 while preserving the full data strength for precise positioning. The resulting single data stream is that of carrier phase measurements with both data noise and bias uncertainty strictly defined. With this mathematical formula the single stream of optimally combined GPS measurements can be efficiently formed by simple numerical calculations. Carrier phase ambiguity resolution, when feasible, is strengthened due to the preserved full data strength with the optimally combined data and the resulting longer wavelength for the ambiguity to be resolved.

Wu, Sien-Chong↗

Computation of optimal low- and medium-thrust orbit transfers

This paper presents the formulation of the optimal low- and medium-thrust orbit transfer control problem, numerical methods for solution, and numerical solutions of the problem. The problem formulation is for final mass maximization and allows for second-harmonic oblateness, atmospheric drag, and 3D noncoplanar nonaligned elliptic terminal orbits. We set up examples to demonstrate the ability of two indirect methods to solve the resulting two point boundary value problems (TPBVP). The methods demonstrated are the multiple point shooting method as formulated in Oberle's (1987) subroutine BOUNDSCO, and the minimizing boundary-condition method (MBCM). We find that although both methods can converge solutions, there are tradeoffs to using either method. We present numerical solutions of planar transfers in which both the initial orbit exit and final orbit entry points have been optimized. These solutions include two- and three-burn transfers. The methods used show an ability to handle thrust down to at least T/W(o) = O(10 exp -3). They also show similar convergence abilities with or without the oblateness and drag terms. We discuss the issue of maximizing with respect to the final time and provide evidence that implies a local optimum at a maximum final time for a given number of burns.

Chuang, C.-H.↗

Solar Thermal Concept Evaluation

Concentrated solar thermal energy can be utilized in a variety of high temperature applications for both terrestrial and space environments. In each application, knowledge of the collector and absorber's heat exchange interaction is required. To understand this coupled mechanism, various concentrator types and geometries, as well as, their relationship to the physical absorber mechanics were investigated. To conduct experimental tests various parts of a 5,000 watt, thermal concentrator, facility were made and evaluated. This was in anticipation at a larger NASA facility proposed for construction. Although much of the work centered on solar thermal propulsion for an upper stage (less than one pound thrust range), the information generated and the facility's capabilities are applicable to material processing, power generation and similar uses. The numerical calculations used to design the laboratory mirror and the procedure for evaluating other solar collectors are presented here. The mirror design is based on a hexagonal faceted system, which uses a spherical approximation to the parabolic surface. The work began with a few two dimensional estimates and continued with a full, three dimensional, numerical algorithm written in FORTRAN code. This was compared to a full geometry, ray trace program, BEAM 4, which optimizes the curvatures, based on purely optical considerations. Founded on numerical results, the characteristics of a faceted concentrator were construed. The numerical methodologies themselves were evaluated and categorized. As a result, the three-dimensional FORTRAN code was the method chosen to construct the mirrors, due to its overall accuracy and superior results to the ray trace program. This information is being used to fabricate and subsequently, laser map the actual mirror surfaces. Evaluation of concentrator mirrors, thermal applications and scaling the results of the 10 foot diameter mirror to a much larger concentrator, were studied. Evaluations, recommendations and pit falls regarding the structure, materials and facility design are presented.

Hawk, Clark W.↗

Gradient-Based Aeroservoelastic Optimization with Static Output Feedback

Static output feedback considers an optimal low-order feedback matrix which directly connects the sensors to the control inputs. This work demonstrates the numerical techniques needed to compute the analytical gradient of the optimal feedback matrix with respect to design variables, which may then be used for gradient-based optimization. The derivatives are demonstrated for aeroservoelastic optimization under a series of closed- loop gust load alleviation constraints, considering a continuous stochastic gust load applied to a transport vehicle configuration, among other design constraints such as utter and maneuver loads. The optimal trade-o s between passive load alleviation and active load alleviation for static output feedback are compared with those from full-state feedback, which may be considered an upper-bound for effective sensor-based control.

Stanford, Bret K.↗

Level-set topology optimization with PDE generated conformal meshes

This paper presents a level-set topology optimization approach that uses conformal meshes for the analysis of the displacement field. The structure’s boundary is represented by the iso-contour of a level-set field discretized on a fixed background design mesh. The conformal mesh is updated for each design iteration via a PDE based mesh morphing process that identifies the set of facets in the background mesh that are homeomorphic to the boundary and relaxes the homeomorphic mesh to conform to the structure’s boundary and ensure high element quality. The conformal mesh allows for a more accurate computation of the response versus density and some level-set based methods which interpolate material properties using the volume fraction. Numerical examples illustrate the proposed approach by optimizing linear-elastic two- and three-dimensional structures, wherein insight into the performance of the mesh morphing process is provided. The examples also highlight the scalability of the approach.

42 ENGINEERING↗

VAN-DAMME: GPU-accelerated and symmetry-assisted quantum optimal control of multi-qubit systems

We present an open-source software package, VAN-DAMME (Versatile Approaches to Numerically Design, Accelerate, and Manipulate Magnetic Excitations), for massively-parallelized quantum optimal control (QOC) calculations of multi-qubit systems. To enable large QOC calculations, the VAN-DAMME software package utilizes symmetry-based techniques with custom GPU-enhanced algorithms. This combined approach allows for the simultaneous computation of hundreds of matrix exponential propagators that efficiently leverage the intra-GPU parallelism found in high-performance GPUs. In addition, to maximize the computational efficiency of the VAN-DAMME code, we carried out several extensive tests on data layout, computational complexity, memory requirements, and performance. These extensive analyses allowed us to develop computationally efficient approaches for evaluating complex-valued matrix exponential propagators based on Padé approximants. To assess the computational performance of our GPU-accelerated VAN-DAMME code, we carried out QOC calculations of systems containing 10 - 15 qubits, which showed that our GPU implementation is 18.4× faster than the corresponding CPU implementation. Our GPU-accelerated enhancements allow efficient calculations of multi-qubit systems, which can be used for the efficient implementation of QOC applications across multiple domains.

97 MATHEMATICS AND COMPUTING↗

Practical CO2—WAG Field Operational Designs Using Hybrid Numerical-Machine-Learning Approaches

Machine-learning technologies have exhibited robust competences in solving many petroleum engineering problems. The accurate predictivity and fast computational speed enable a large volume of time-consuming engineering processes such as history-matching and field development optimization. The Southwest Regional Partnership on Carbon Sequestration (SWP) project desires rigorous history-matching and multi-objective optimization processes, which fits the superiorities of the machine-learning approaches. Although the machine-learning proxy models are trained and validated before imposing to solve practical problems, the error margin would essentially introduce uncertainties to the results. In this paper, a hybrid numerical machine-learning workflow solving various optimization problems is presented. By coupling the expert machine-learning proxies with a global optimizer, the workflow successfully solves the history-matching and CO2 water alternative gas (WAG) design problem with low computational overheads. The history-matching work considers the heterogeneities of multiphase relative characteristics, and the CO2-WAG injection design takes multiple techno-economic objective functions into accounts. This work trained an expert response surface, a support vector machine, and a multi-layer neural network as proxy models to effectively learn the high-dimensional nonlinear data structure. The proposed workflow suggests revisiting the high-fidelity numerical simulator for validation purposes. The experience gained from this work would provide valuable guiding insights to similar CO2 enhanced oil recovery (EOR) projects.

29 ENERGY PLANNING, POLICY, AND ECONOMY↗

Numerical and experimental calibration of calorimetric sample cell dedicated to nuclear heating measurements

Online nuclear measurements inside experimental channels of MTRs are needed for experimental works (to design set-ups) and for numerical works (input data) in order to better understanding complex phenomena occurring during the accelerated ageing of materials and the irradiation of nuclear fuel. In this paper we focus only on one kind of measurements: nuclear heating performed by means of a radiometric calorimeter. The aims of numerical and experimental works are firstly to optimize the sensor response for new energy deposit ranges and then to miniaturize this sensor for JHR irradiation conditions A first calorimeter, developed previously by the CEA, is studied. It corresponds to a graphite differential calorimeter divided into two twin cells (a reference cell, and a sample one). It is used with a non adiabatic mode or heat flow mode. Experimental calibration of the sample cell is presented. In that case, energy deposit is simulated by Joule effect and the sample cell is inserted into a bath at a regulated temperature and controlled flow. The response of the sensor is shown versus electrical power imposed for two flow regimes (intensive or moderated forced convection). These experimental results are compared to numerical works and improvements are discussed. (authors)

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗

Singular-Arc Time-Optimal Trajectory of Aircraft in Two-Dimensional Wind Field

This paper presents a study of a minimum time-to-climb trajectory analysis for aircraft flying in a two-dimensional altitude dependent wind field. The time optimal control problem possesses a singular control structure when the lift coefficient is taken as a control variable. A singular arc analysis is performed to obtain an optimal control solution on the singular arc. Using a time-scale separation with the flight path angle treated as a fast state, the dimensionality of the optimal control solution is reduced by eliminating the lift coefficient control. A further singular arc analysis is used to decompose the original optimal control solution into the flight path angle solution and a trajectory solution as a function of the airspeed and altitude. The optimal control solutions for the initial and final climb segments are computed using a shooting method with known starting values on the singular arc The numerical results of the shooting method show that the optimal flight path angle on the initial and final climb segments are constant. The analytical approach provides a rapid means for analyzing a time optimal trajectory for aircraft performance.

Nguyen, Nhan↗

Thermal Conductivity Estimation from Transient Test Data with Embedded Thermocouples using Genetic Algorithm Optimization

Inverse heat transfer methodology previously developed to estimate thermal properties of high temperature fibrous insulation from embedded thermocouples was applied to the thermoplastic polymer, polyether-ether-ketone (PEEK). A small experimental setup was utilized to test the PEEK sample between temperatures of 300 K and 525 K at atmospheric pressure. Cylindrical plugs of PEEK with three thermocouples embedded at various depths were incorporated in the test sample. The experimental PEEK thermocouple data were used as the boundary and initial conditions of a one-dimensional numerical thermal model to predict the internal temperatures of the material. The thermal conductivity of PEEK was estimated with the Continuous Genetic Algorithm optimization technique by searching for the coefficients of a functional form of thermal conductivity that minimized the difference between the experimentally measured and model predicted internal temperature values. The thermal testing, one-dimensional numerical thermal model, optimization algorithm, analysis, and results are presented.

Thermal Properties↗

Optimal Design of Passive Flow Control for a Boundary-Layer-Ingesting Offset Inlet Using Design-of-Experiments

This research will investigate the use of Design-of-Experiments (DOE) in the development of an optimal passive flow control vane design for a boundary-layer-ingesting (BLI) offset inlet in transonic flow. This inlet flow control is designed to minimize the engine fan face distortion levels and first five Fourier harmonic half amplitudes while maximizing the inlet pressure recovery. Numerical simulations of the BLI inlet are computed using the Reynolds-averaged Navier-Stokes (RANS) flow solver, OVERFLOW, developed at NASA. These simulations are used to generate the numerical experiments for the DOE response surface model. In this investigation, two DOE optimizations were performed using a D-Optimal Response Surface model. The first DOE optimization was performed using four design factors which were vane height and angles-of-attack for two groups of vanes. One group of vanes was placed at the bottom of the inlet and a second group symmetrically on the sides. The DOE design was performed for a BLI inlet with a free-stream Mach number of 0.85 and a Reynolds number of 2 million, based on the length of the fan face diameter, matching an experimental wind tunnel BLI inlet test. The first DOE optimization required a fifth order model having 173 numerical simulation experiments and was able to reduce the DC60 baseline distortion from 64% down to 4.4%, while holding the pressure recovery constant. A second DOE optimization was performed holding the vanes heights at a constant value from the first DOE optimization with the two vane angles-of-attack as design factors. This DOE only required a second order model fit with 15 numerical simulation experiments and reduced DC60 to 3.5% with small decreases in the fourth and fifth harmonic amplitudes. The second optimal vane design was tested at the NASA Langley 0.3-Meter Transonic Cryogenic Tunnel in a BLI inlet experiment. The experimental results showed a 80% reduction of DPCPavg, the circumferential distortion level at the engine fan face.

Allan, Brian G.↗

Zgoubi Users' Guide

The computer code zgoubi calculates trajectories of charged particles in magnetic and electric fields. At the origin specially adapted to the definition and adjustment of beam lines and magnetic spectrometers, it has so evolved that it allows the study of systems including complex sequences of optical elements such as dipoles, quadrupoles, arbitrary multipoles, FFAG magnets and other magnetic or electric devices, and is able as well to handle periodic structures. Compared to other codes, it presents several peculiarities, as follows - a non-exhaustive list : • a numerical method for integrating the Lorentz equation, based on Taylor series, which optimizes computing time and provides high accuracy and strong symplecticity, • spin tracking, using the same numerical method as for the Lorentz equation, • account of stochastic photon emission, and its effects on particle dynamics, • calculation of the synchrotron radiation electric field and spectra in arbitrary magnetic fields, from the ray-tracing outcomes, • the possibility of using a mesh, which allows ray-tracing from simulated or measured (1-D, 2-D or 3-D) electric and magnetic field maps, • a number Monte Carlo procedures : unlimited number of trajectories, in-flight decay, stochastic radiation, etc., • built-in fitting procedures allowing arbitrary variables and a large variety of constraints, easily expandable, • multi-turn tracking in circular accelerators including features proper to machine parameter calculation and survey, • simulation of time-varying power supplies, and • simulation of arbitrary radio-frequency programs. The initial version of zgoubi was dedicated to ray-tracing in magnetic elements, beam lines, spectrometers. It was perfected for the purpose of studying, and operating, the four spectrometers SPES I, II, III, IV at the Laboratoire National SATURNE (CEA-Saclay, France), and, later, SPEG at GANIL (Caen, France). Developments regarding spin tracking and ray-tracing in combined electric and magnetic fields were implemented, in the late 1980s and early 1990s respectively. In the mid-1990s, the computation of synchrotron radiation electromagnetic impulse and spectra was introduced, for the purpose of synchrotron radiation diagnostic R&D at LEP, and further applied to the design of the SR diagnostics installations at LHC in the early 2000s. In the mean time, several new optical elements were added, such as electro-magnetic and other electrostatic lenses. Used since several years for special studies in periodic machines (e.g., SATURNE at Saclay, COSY at Julich, LEP and LHC at CERN), zgoubi has also undergone extensive developments regarding storage ring related features. Many developments have been accomplished since the early 2000s in the frame of a number of project design and beam dynamics studies, as the neutrino factory, lepton and hadron colliders, spin studies at AGS and RHIC, etc. As a consequence the list of optical elements and the compendium of numerical methods, so-called “Glossary of Keywords” list, pp. 7 and 169, has stretched with new simulation and computing procedures, ranging from fitting (FIT2 procedure ; additional constraints) to overlapping magnetic field capabilities (DIPOLES, FFAG), spin manipulation (SPINR), optical elements (BEAMBEAM), radiation damping tools (SRLOSS) and many others (FAISTORE, MAP2D-E, OPTICS, PICKUPS, TWISS, etc.). The graphic interface to zgoubi (zpop, Part D) has undergone extensive developments, making it a convenient companion tool to the use of zgoubi.

43 PARTICLE ACCELERATORS↗

Prediction of pilot opinion ratings using an optimal pilot model

A brief review of some of the more pertinent applications of analytical pilot models to the prediction of aircraft handling qualities is undertaken. The relative ease with which multiloop piloting tasks can be modeled via the optimal control formulation makes the use of optimal pilot models particularly attractive for handling qualities research. To this end, a rating hypothesis is introduced which relates the numerical pilot opinion rating assigned to a particular vehicle and task to the numerical value of the index of performance resulting from an optimal pilot modeling procedure as applied to that vehicle and task. This hypothesis is tested using data from piloted simulations and is shown to be reasonable. An example concerning a helicopter landing approach is introduced to outline the predictive capability of the rating hypothesis in multiaxis piloting tasks.

Hess, R. A.↗