Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “optimization methods”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

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

At least 109 records · Page 6
All sourcesNASA NTRSDOE OSTI
DOE OSTI2019

Design optimization methods for high-performance research reactor core design

Typical objectives of high-performance research reactors include neutron scattering, materials irradiation, and isotope production. The creation of a core that optimizes these multiple objectives often presents a multidimensional (i.e., in design space), multiobjective optimization problem. The developed systematic approach discussed herein draws on response surface methodology to validate and leverage a surrogate model to serve in place of high-fidelity computational analyses. Optimization and design analysis methods leverage this surrogate model to provide a flexible tool for generating optimized designs and understanding the impact of design decisions on desired metrics. In applications to High Flux Isotope Reactor (HFIR) low-enriched uranium (LEU) core designs, neutronic, isotopic evolution, and thermal hydraulic analyses are used to generate key performance and safety metrics for assessing the feasibility and fitness of given designs. Three optimized designs that determine different desired metrics and constraints (e.g., key metric weighting and fabrication constraints) are presented, providing potential design options that satisfy the requirements for HFIR’s conversion from high enriched uranium to LEU fuel.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗
DOE OSTI2014

TU-F-BRE-03: Application of a Novel Mass-Density Compensation Optimization Method to Improve the Response of a Liquid-Filled Ionization Chamber in Nonstandard Fields

Purpose: To improve the response of the microLion detector (PTW 31018) in small field conditions by optimizing the density of the detector's non-sensitive components using a novel method based on the detector's dose response function. Methods: The central values h(0,0) of the perturbation functions for the microLion detector and a volume of water equivalent to its sensitive volume were calculated using the Monte Carlo user code egs-chamber by scoring the dose absorbed by a 6 MV photon pencil beam incident on their centroids. Values of h(0,0) were plotted as a function of the density of the microLion's graphite electrode with the detector placed in the axial orientation. The optimized density was found by finding the minimal value of h(0,0). Results: A density of 1.37 g/cm3 was found to minimize the perturbation function of the microLion detector. The modified microLion's response was then evaluated in small square fields with sides in the range of 5 – 40 mm and found to be consistent with highly watere-quivalent detectors such as a scintillating detector (Exradin W1) and a generic alanine detector in both axial and radial orientations. Conclusion: This work illustrates a novel method which can used to optimize the design of radiation detectors in small fields. This method should also work with other optimization parameters (e.g. thickness of electrode). Density-compensated detectors have the potential to eliminate the need to evaluate nonstandard field correction factors as described by the IAEA-AAPM formalism (Alfonso et al.) and simplify future dosimetry protocols for SRS/SBRT modalities. Finally, we also expect an improvement in the response of density-compensated detectors for composite IMRT fields.

07 ISOTOPE AND RADIATION SOURCES↗
NASA NTRS1992

Flow analysis and design optimization methods for nozzle-afterbody of a hypersonic vehicle

This report summarizes the methods developed for the aerodynamic analysis and the shape optimization of the nozzle-afterbody section of a hypersonic vehicle. Initially, exhaust gases were assumed to be air. Internal-external flows around a single scramjet module were analyzed by solving the 3D Navier-Stokes equations. Then, exhaust gases were simulated by a cold mixture of Freon and Ar. Two different models were used to compute these multispecies flows as they mixed with the hypersonic airflow. Surface and off-surface properties were successfully compared with the experimental data. The Aerodynamic Design Optimization with Sensitivity analysis was then developed. Pre- and postoptimization sensitivity coefficients were derived and used in this quasi-analytical method. These coefficients were also used to predict inexpensively the flow field around a changed shape when the flow field of an unchanged shape was given. Starting with totally arbitrary initial afterbody shapes, independent computations were converged to the same optimum shape, which rendered the maximum axial thrust.

Baysal, O.↗
NASA NTRS1974

On the application of deterministic optimization methods to stochastic control problems

A technique is presented by which deterministic optimization techniques, for example, the maximum principle of Pontriagin, can be applied to stochastic optimal control problems formulated around linear systems with Gaussian noises and general cost criteria. Using this technique, the stochastic nature of the problem is suppressed but for two expectation operations, the optimization being deterministic. The use of the technique in treating problems with quadratic and nonquadratic costs is illustrated.

Kramer, L. C.↗
NASA NTRS2002

A Comparison of Trajectory Optimization Methods for the Impulsive Minimum Fuel Rendezvous Problem

In this paper we present a comparison of optimization approaches to the minimum fuel rendezvous problem. Both indirect and direct methods are compared for a variety of test cases. The indirect approach is based on primer vector theory. The direct approaches are implemented numerically and include Sequential Quadratic Programming (SQP), Quasi-Newton, Simplex, Genetic Algorithms, and Simulated Annealing. Each method is applied to a variety of test cases including, circular to circular coplanar orbits, LEO to GEO, and orbit phasing in highly elliptic orbits. We also compare different constrained optimization routines on complex orbit rendezvous problems with complicated, highly nonlinear constraints.

Hughes, Steven P.↗
DOE OSTI2024

Two-Step Hyperparameter Optimization Method: Accelerating Hyperparameter Search by Using a Fraction of a Training Dataset

Abstract Hyperparameter optimization (HPO) is an important step in machine learning (ML) model development, but common practices are archaic—primarily relying on manual or grid searches. This is partly because adopting advanced HPO algorithms introduces added complexity to the workflow, leading to longer computation times. This poses a notable challenge to ML applications, as suboptimal hyperparameter selections curtail the potential of ML model performance, ultimately obstructing the full exploitation of ML techniques. In this article, we present a two-step HPO method as a strategic solution to curbing computational demands and wait times, gleaned from practical experiences in applied ML parameterization work. The initial phase involves a preliminary evaluation of hyperparameters on a small subset of the training dataset, followed by a reevaluation of the top-performing candidate models postretraining with the entire training dataset. This two-step HPO method is universally applicable across HPO search algorithms, and we argue it has attractive efficiency gains. As a case study, we present our recent application of the two-step HPO method to the development of neural network emulators for aerosol activation. Although our primary use case is a data-rich limit with many millions of samples, we also find that using up to 0.0025% of the data—a few thousand samples—in the initial step is sufficient to find optimal hyperparameter configurations from much more extensive sampling, achieving up to 135× speedup. The benefits of this method materialize through an assessment of hyperparameters and model performance, revealing the minimal model complexity required to achieve the best performance. The assortment of top-performing models harvested from the HPO process allows us to choose a high-performing model with a low inference cost for efficient use in global climate models (GCMs).

97 MATHEMATICS AND COMPUTING↗
NASA NTRS1986

Optimization methods and silicon solar cell numerical models

The goal of this project is the development of an optimization algorithm for use with a solar cell model. It is possible to simultaneously vary design variables such as impurity concentrations, front junction depth, back junctions depth, and cell thickness to maximize the predicted cell efficiency. An optimization algorithm has been developed and interfaced with the Solar Cell Analysis Program in 1 Dimension (SCAPID). SCAPID uses finite difference methods to solve the differential equations which, along with several relations from the physics of semiconductors, describe mathematically the operation of a solar cell. A major obstacle is that the numerical methods used in SCAPID require a significant amount of computer time, and during an optimization the model is called iteratively until the design variables converge to the value associated with the maximum efficiency. This problem has been alleviated by designing an optimization code specifically for use with numerically intensive simulations, to reduce the number of times the efficiency has to be calculated to achieve convergence to the optimal solution. Adapting SCAPID so that it could be called iteratively by the optimization code provided another means of reducing the cpu time required to complete an optimization. Instead of calculating the entire I-V curve, as is usually done in SCAPID, only the efficiency is calculated (maximum power voltage and current) and the solution from previous calculations is used to initiate the next solution.

Girardini, K.↗
NASA NTRS2003

A Comparison of Trajectory Optimization Methods for the Impulsive Minimum Fuel Rendezvous Problem

In this paper we present, a comparison of trajectory optimization approaches for the minimum fuel rendezvous problem. Both indirect and direct methods are compared for a variety of test cases. The indirect approach is based on primer vector theory. The direct approaches are implemented numerically and include Sequential Quadratic Programming (SQP). Quasi- Newton and Nelder-Meade Simplex. Several cost function parameterizations are considered for the direct approach. We choose one direct approach that appears to be the most flexible. Both the direct and indirect methods are applied to a variety of test cases which are chosen to demonstrate the performance of each method in different flight regimes. The first test case is a simple circular-to-circular coplanar rendezvous. The second test case is an elliptic-to-elliptic line of apsides rotation. The final test case is an orbit phasing maneuver sequence in a highly elliptic orbit. For each test case we present a comparison of the performance of all methods we consider in this paper.

Hughes, Steven P.↗
NASA NTRS1987

Optimization methods applied to the aerodynamic design of helicopter rotor blades

Described is a formal optimization procedure for helicopter rotor blade design which minimizes hover horsepower while assuring satisfactory forward flight performance. The approach is to couple hover and forward flight analysis programs with a general-purpose optimization procedure. The resulting optimization system provides a systematic evaluation of the rotor blade design variables and their interaction, thus reducing the time and cost of designing advanced rotor blades. The paper discusses the basis for and details of the overall procedure, describes the generation of advanced blade designs for representative Army helicopters, and compares design and design effort with those from the conventional approach which is based on parametric studies and extensive cross-plots.

Walsh, Joanne L.↗
NASA NTRS1985

Optimization methods applied to the aerodynamic design of helicopter rotor blades

This paper describes a formal optimization procedure for helicopter rotor blade designs which minimizes hover horsepower while assuring satisfactory forward flight performance. The approach is to couple hover and forward flight analysis programs with a general purpose optimization procedure. The resulting optimization system provides a systematic evaluation of the rotor blade design variables and their interaction, thus reducing the time and cost of designing advanced rotor blades. The paper discusses the basis for and details of the overall procedure, describes the generation of advanced blade designs for representative Army helicopters, and compares designs and design effort with those from the conventional approach which is based on parametric studies and extensive cross-plots.

Walsh, J. L.↗
NASA NTRS1992

Supersonic airplane design optimization method for aerodynamic performance and low sonic boom

This paper presents a new methodology for the optimization of supersonic airplane designs to meet the dual design objectives of low sonic boom and high aerodynamic performance. Two sets of design parameters are used on an existing High Speed Civil Transport (HSCT) configuration to maximize the aerodynamic performance and minimize the sonic boom under the flight track. One set of the parameters perturbs the camber line of the wing sections to maximize the lift-over-drag ratio (L/D). A preliminary optimization run yielded a 3.75 percent improvement in L/D over a baseline low-boom configuration. The other set of parameters modifies the fuselage area to achieve a target F-function. Starting from an initial configuration with strong bow, wing, and tail shocks, a modified design with a flat-top signature is obtained. The methods presented can easily incorporate other design variables and objective functions. Extensions to the present capability in progress are described.

Cheung, Samson H.↗
NASA NTRS1990

Global optimization methods for engineering design

The problem is to find a global minimum for the Problem P. Necessary and sufficient conditions are available for local optimality. However, global solution can be assured only under the assumption of convexity of the problem. If the constraint set S is compact and the cost function is continuous on it, existence of a global minimum is guaranteed. However, in view of the fact that no global optimality conditions are available, a global solution can be found only by an exhaustive search to satisfy Inequality. The exhaustive search can be organized in such a way that the entire design space need not be searched for the solution. This way the computational burden is reduced somewhat. It is concluded that zooming algorithm for global optimizations appears to be a good alternative to stochastic methods. More testing is needed; a general, robust, and efficient local minimizer is required. IDESIGN was used in all numerical calculations which is based on a sequential quadratic programming algorithm, and since feasible set keeps on shrinking, a good algorithm to find an initial feasible point is required. Such algorithms need to be developed and evaluated.

Arora, Jasbir S.↗
DOE OSTI2023

NSGA-PINN: A Multi-Objective Optimization Method for Physics-Informed Neural Network Training

This paper presents NSGA-PINN, a multi-objective optimization framework for the effective training of physics-informed neural networks (PINNs). The proposed framework uses the non-dominated sorting genetic algorithm (NSGA-II) to enable traditional stochastic gradient optimization algorithms (e.g., ADAM) to escape local minima effectively. Additionally, the NSGA-II algorithm enables satisfying the initial and boundary conditions encoded into the loss function during physics-informed training precisely. We demonstrate the effectiveness of our framework by applying NSGA-PINN to several ordinary and partial differential equation problems. In particular, we show that the proposed framework can handle challenging inverse problems with noisy data.

Lu, Binghang (ORCID:0009000160016632)↗
DOE OSTI2018

Development Of A Novel Genetic Algorithm-based Optimization Method For Design Of Nuclear Criticality Experiments

A new evaluation of isotopic molybdenum total neutron cross section data based on measurements performed at Rensselaer Polytechnic Institute prompted researchers at Rensselaer and Los Alamos National Laboratory (LANL) to design a nuclear criticality experiment for the validation of the data. A novel approach of criticality experiment design employing a genetic algorithm-based optimization technique was developed. The method produced a critical experiment with a cross section sensitivity that exceeded target specifications by an order of magnitude.

59 BASIC BIOLOGICAL SCIENCES↗
NASA NTRS1990

Genetic-evolution-based optimization methods for engineering design

This paper presents the applicability of a biological model, based on genetic evolution, for engineering design optimization. Algorithms embodying the ideas of reproduction, crossover, and mutation are developed and applied to solve different types of structural optimization problems. Both continuous and discrete variable optimization problems are solved. A two-bay truss for maximum fundamental frequency is considered to demonstrate the continuous variable case. The selection of locations of actuators in an actively controlled structure, for minimum energy dissipation, is considered to illustrate the discrete variable case.

Rao, S. S.↗
DOE OSTI2020

Total variation vs L1 regularization: a comparison of compressive sensing optimization methods for chemical detection

One of the fundamental assumptions of compressive sensing (CS) is that a signal can be reconstructed from a small number of samples by solving an optimization problem with the appropriate regularization term. Two standard regularization terms are the L1 norm and the total variation (TV) norm. We present a comparison of CS reconstruction results based on these two approaches in the context of chemical detection, and we demonstrate that optimization based on the L1 norm outperforms optimization based on the TV norm. Our comparison is driven by CS sampling, reconstruction, and chemical detection in two real-world datasets: the Physical Sciences Inc. Fabry-Perot interferometer sensor multispectral dataset and the Johns Hopkins Applied Physics Lab FTIR-based longwave infrared sensor hyperspectral dataset. Both datasets contain the release of a chemical simulant such as glacial acetic acid, triethyl phosphate, and sulfur hexafluoride. For chemical detection we use the adaptive coherence estimator (ACE) and bulk coherence, and we propose algorithmic ACE thresholds to define the presence or absence of a chemical of interest in both un-compressed data cubes and reconstructed data cubes. The un-compressed data cubes provide an approximate ground truth. We demonstrate that optimization based on either the L1 norm or TV norm results in successful chemical detection at a compression rate of 90%, but we show that L1 optimization is preferable. We present quantitative comparisons of chemical detection on reconstructions from the two methods, with an emphasis on the number of pixels with an ACE value above the threshold.

Hyperspectral imagery, L1-norm, total variation, A↗
NASA NTRS1972

On the application of deterministic optimization methods to stochastic control problems.

A technique is presented by which one can apply the Minimum Principle of Pontryagin to stochastic optimal control problems formulated around linear systems with Gaussian noises and general cost criteria. Using this technique, the stochastic nature of the problem is suppressed but for two expectation operations, the optimization being essentially deterministic. The technique is applied to systems with quadratic and non-quadratic costs to illustrate its use.

Kramer, L. C.↗