Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “code 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 883 records · Page 49

Rotorcraft airframe structural optimization for combined vibration and fatigue constraints

This paper addresses the application of a formal optimization technique in rotorcraft airframe structural design studies to reduce the structural weight, to lower airframe vibrations, and to enhance fatigue life of the structure. Vibration and fatigue considerations in airframe design are described. An optimization methodology based on the use of a nonlinear programming technique to size airframe structural members subjected to constraints on weight, vibration response and fatigue stresses under dynamic loads, are described. The paper focuses on the development of necessary computational tools for airframe structural optimization and describes the sensitivity analysis procedure for these types of design constraints. Further, the paper describes the optimization procedure as implemented in a computer code called DYNOPT which is a unique operational combination of several newly developed Fortran codes as well as modification of existing codes consisting of the direct matrix abstraction modules of the MSC/NASTRAN Program and CONMIN optimizer. The application of the optimization procedure is demonstrated using an elastic-line model of the Bell AH-1G helicopter airframe structure and computational results are discussed.

Sareen, Ashish K.↗

Systematic Propulsion Optimization Tools (SPOT)

This paper describes a computer program written by senior-level Mechanical Engineering students at the University of Alabama in Huntsville which is capable of optimizing user-defined delivery systems for carrying payloads into orbit. The custom propulsion system is designed by the user through the input of configuration, payload, and orbital parameters. The primary advantages of the software, called Systematic Propulsion Optimization Tools (SPOT), are a user-friendly interface and a modular FORTRAN 77 code designed for ease of modification. The optimization of variables in an orbital delivery system is of critical concern in the propulsion environment. The mass of the overall system must be minimized within the maximum stress, force, and pressure constraints. SPOT utilizes the Design Optimization Tools (DOT) program for the optimization techniques. The SPOT program is divided into a main program and five modules: aerodynamic losses, orbital parameters, liquid engines, solid engines, and nozzles. The program is designed to be upgraded easily and expanded to meet specific user needs. A user's manual and a programmer's manual are currently being developed to facilitate implementation and modification.

Bower, Mark↗

Improving turbo-like codes using iterative decoder analysis

The density evolution method is used to analyze the performance and optimize the structure of parallel and serial turbo codes, and generalized serial concatenations of mixtures of different outer and inner codes. Design examples are given for mixture codes.

turbo codes iterative decoding↗

Maximum Likelihood Expectation-Maximization Algorithms Applied to Localization and Identification of Radioactive Sources with Recent Coded Mask Gamma Cameras

In this work, we tested Maximum likelihood expectation-maximization (MLEM) algorithms optimized for gamma imaging applications on two recent coded mask gamma cameras. We respectively took advantage of the characteristics of the GAMPIX and Caliste HD-based gamma cameras: noise reduction thanks to mask/anti-mask procedure but limited energy resolution for GAMPIX, high energy resolution for Caliste HD. One of our short-term perspectives is the test of MAPEM algorithms integrating specific prior values for the data to reconstruct adapted to the gamma imaging topic. (authors)

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Constrained Bayesian Optimization of Criticality Experiments [Slides]

The design of criticality experiments is typically an iterative process that employs a Monte Carlo transport code. The goal is to find a design that optimizes some variable, like the sensitivity of a response to a cross section, while simultaneously ensuring criticality. The high fidelity of the Monte Carlo code is a great asset, but it makes exploring the design space computationally expensive. Herein, we present how a constrained Bayesian optimization algorithm can be used to efficiently design a criticality experiment. It uses Gaussian processes as a surrogate model to probe the design space and to reduce the number of code executions that are needed to find the optimum. We demonstrate constrained Bayesian optimization with a Pu-239/polyethylene solution system and a TEX experiment that is designed for criticality safety validation of a nuclear waste model at the Hanford Site. For both systems, a global optimum was found within 75 Monte Carlo simulations.

42 ENGINEERING↗

Necessary conditions for the optimality of variable rate residual vector quantizers

Residual vector quantization (RVQ), or multistage VQ, as it is also called, has recently been shown to be a competitive technique for data compression. The competitive performance of RVQ reported in results from the joint optimization of variable rate encoding and RVQ direct-sum code books. In this paper, necessary conditions for the optimality of variable rate RVQ's are derived, and an iterative descent algorithm based on a Lagrangian formulation is introduced for designing RVQ's having minimum average distortion subject to an entropy constraint. Simulation results for these entropy-constrained RVQ's (EC-RVQ's) are presented for memory less Gaussian, Laplacian, and uniform sources. A Gauss-Markov source is also considered. The performance is superior to that of entropy-constrained scalar quantizers (EC-SQ's) and practical entropy-constrained vector quantizers (EC-VQ's), and is competitive with that of some of the best source coding techniques that have appeared in the literature.

Kossentini, Faouzi↗

Development of an hp-version finite element method for computational optimal control

The purpose of this research effort was to begin the study of the application of hp-version finite elements to the numerical solution of optimal control problems. Under NAG-939, the hybrid MACSYMA/FORTRAN code GENCODE was developed which utilized h-version finite elements to successfully approximate solutions to a wide class of optimal control problems. In that code the means for improvement of the solution was the refinement of the time-discretization mesh. With the extension to hp-version finite elements, the degrees of freedom include both nodal values and extra interior values associated with the unknown states, co-states, and controls, the number of which depends on the order of the shape functions in each element. One possible drawback is the increased computational effort within each element required in implementing hp-version finite elements. We are trying to determine whether this computational effort is sufficiently offset by the reduction in the number of time elements used and improved Newton-Raphson convergence so as to be useful in solving optimal control problems in real time. Because certain of the element interior unknowns can be eliminated at the element level by solving a small set of nonlinear algebraic equations in which the nodal values are taken as given, the scheme may turn out to be especially powerful in a parallel computing environment. A different processor could be assigned to each element. The number of processors, strictly speaking, is not required to be any larger than the number of sub-regions which are free of discontinuities of any kind.

Hodges, Dewey H.↗

Verification of Reduced Order Modeling Based Uncertainty/Sensitivity Estimator (ROMUSE)

This article presents a number of verification case studies for a recently developed sensitivity/uncertainty code package. The code package, ROMUSE (Reduced Order Modeling based Uncertainty/Sensitivity Estimator) is an effort to provide an analysis tool to be used in conjunction with reactor core simulators, in particular the Virtual Environment for Reactor Applications (VERA) core simulator. ROMUSE has been written in C++ and is currently capable of performing various types of parameter perturbations and associated sensitivity analysis, uncertainty quantification, surrogate model construction and subspace analysis. The current version 2.0 has the capability to interface with the Design Analysis Kit for Optimization and Terascale Applications (DAKOTA) code, which gives ROMUSE access to the various algorithms implemented within DAKOTA, most importantly model calibration. The verification study is performed via two basic problems and two reactor physics models. The first problem is used to verify the ROMUSE single physics gradient-based range finding algorithm capability using an abstract quadratic model. The second problem is the Brusselator problem, which is a coupled problem representative of multi-physics problems. This problem is used to test the capability of constructing surrogates via ROMUSE-DAKOTA. Finally, light water reactor pin cell and sodium-cooled fast reactor fuel assembly problems are simulated via SCALE 6.1 to test ROMUSE capability for uncertainty quantification and sensitivity analysis purposes.

97 MATHEMATICS AND COMPUTING↗

AutoBayes Program Synthesis System Users Manual

Program synthesis is the systematic, automatic construction of efficient executable code from high-level declarative specifications. AutoBayes is a fully automatic program synthesis system for the statistical data analysis domain; in particular, it solves parameter estimation problems. It has seen many successful applications at NASA and is currently being used, for example, to analyze simulation results for Orion. The input to AutoBayes is a concise description of a data analysis problem composed of a parameterized statistical model and a goal that is a probability term involving parameters and input data. The output is optimized and fully documented C/C++ code computing the values for those parameters that maximize the probability term. AutoBayes can solve many subproblems symbolically rather than having to rely on numeric approximation algorithms, thus yielding effective, efficient, and compact code. Statistical analysis is faster and more reliable, because effort can be focused on model development and validation rather than manual development of solution algorithms and code.

Schumann, Johann↗

A technique for integrating engine cycle and aircraft configuration optimization

A method for conceptual aircraft design that incorporates the optimization of major engine design variables for a variety of cycle types was developed. The methodology should improve the lengthy screening process currently involved in selecting an appropriate engine cycle for a given application or mission. The new capability will allow environmental concerns such as airport noise and emissions to be addressed early in the design process. The ability to rapidly perform optimization and parametric variations using both engine cycle and aircraft design variables, and to see the impact on the aircraft, should provide insight and guidance for more detailed studies. A brief description of the aircraft performance and mission analysis program and the engine cycle analysis program that were used is given. A new method of predicting propulsion system weight and dimensions using thermodynamic cycle data, preliminary design, and semi-empirical techniques is introduced. Propulsion system performance and weights data generated by the program are compared with industry data and data generated using well established codes. The ability of the optimization techniques to locate an optimum is demonstrated and some of the problems that had to be solved to accomplish this are illustrated. Results from the application of the program to the analysis of three supersonic transport concepts installed with mixed flow turbofans are presented. The results from the application to a Mach 2.4, 5000 n.mi. transport indicate that the optimum bypass ratio is near 0.45 with less than 1 percent variation in minimum gross weight for bypass ratios ranging from 0.3 to 0.6. In the final application of the program, a low sonic boom fix a takeoff gross weight concept that would fly at Mach 2.0 overwater and at Mach 1.6 overland is compared with a baseline concept of the same takeoff gross weight that would fly Mach 2.4 overwater and subsonically overland. The results indicate that for the design mission, the low boom concept has a 5 percent total range penalty relative to the baseline. Additional cycles were optimized for various design overland distances and the effect of flying off-design overland distances is illustrated.

Geiselhart, Karl A.↗

A parallel variable population multi-objective optimizer for accelerator beam dynamics optimization

The simultaneous optimization of multiple objective functions is needed in many particle accelerator applications. In this paper, we present a parallel evolution based multi-objective optimizer that uses a variable population from generation to generation and an external storage to save good solutions. Two heuristic optimization methods, one uses the unified differential evolution and the other uses the real-coded genetic algorithm, are included in the optimizer to generate next generation candidate solutions, and are compared in the test examples. Finally, as an application, we applied this optimizer to the beam dynamics design optimization of a photoinjector and attained the optimal front solutions after 200 generations with the unified differential evolution offspring production scheme.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

pLiner

Compiler optimizations can alter significantly the numerical results of scientific computing applications. When numerical results differ significantly between compilers, optimization levels, and floating-point hardware, these numerical inconsistencies can impact programming productivity. pLiner is a framework that helps programmers identify locations in the source code that are highly affected by compiler optimizations. pLiner uses a novel approach to identify such code locations by enhancing the floatingpoint precision of variables and expressions. Using a guided search to locate the most significant code regions, pLiner can report to users such locations at different granularities, file, function, and line of code.

Laguna Peralta, Ignacio↗

Integration of Heat Transfer, Stress, and Particle Trajectory Simulation

Calabazas Creek Research, Inc. developed and currently markets Beam Optics Analyzer (BOA) in the United States and abroad. BOA is a 3D, charged particle optics code that solves the electric and magnetic fields with and without the presence of particles. It includes automatic and adaptive meshing to resolve spatial scales ranging from a few millimeters to meters. It is fully integrated with CAD packages, such as SolidWorks, allowing seamless geometry updates. The code includes iterative procedures for optimization, including a fully functional, graphical user interface. Recently, time dependent, particle in cell capability was added, pushing particles synchronically under quasistatic electromagnetic fields to obtain particle bunching under RF conditions. A heat transfer solver was added during this Phase I program. Completed tasks include: (1) Added a 3D finite element heat transfer solver with adaptivity; (2) Determined the accuracy of the linear heat transfer field solver to provide the basis for development of higher order solvers in Phase II; (3) Provided more accurate and smoother power density fields; and (4) Defined the geometry using the same CAD model, while maintaining different meshes, and interfacing the power density field between the particle simulator and heat transfer solvers. These objectives were achieved using modern programming techniques and algorithms. All programming was in C++ and parallelization in OpenMP, utilizing state-of-the-art multi-core technology. Both x86 and x64 versions are supported. The GUI design and implementation used Microsoft Foundation Class.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

LibraryX: A Framework for Cross-Library-Call Optimization

Scientific applications utilize performance libraries as a software engineering concept: these libraries encapsulate important and well-understood (mathematical) operations, allow for reuse, and are implemented and tuned by experts. Domain scientists then implement complex algorithms based on these domainspecific libraries. While individual library calls are optimized, larger performance gains across sequences of calls—sometimes spanning multiple libraries—are often unrealized, forcing a trade-off between performance and implementation complexity.To overcome this issue, we propose LibraryX, an approach and a system that allows for cross-library-call optimization even when library calls stem from multiple performance libraries. LibraryX annotates library calls with semantic information and optimizes entire directed acyclic graphs (DAGs) of calls dynamically using the SPIRAL code generation system. We demonstrate its effectiveness across a range of memory bound workloads, achieving significant speedups on Nvidia, AMD, and Intel accelerators compared to code using native libraries without cross-call optimization.

Rao, Sanil [Carnegie Mellon University,Department ↗

Constrained Bayesian Optimization of Criticality Experiments at LLNL [Slides]

The design of criticality experiments is typically an iterative process that employs a Monte Carlo transport code. The goal is to find a design that optimizes some variable, like the sensitivity of a response to a cross section, while simultaneously ensuring criticality. The high fidelity of the Monte Carlo code is a great asset, but it makes exploring the design space computationally expensive. Herein, we present how a constrained Bayesian optimization algorithm can be used to efficiently design a criticality experiment. It uses Gaussian processes as a surrogate model to probe the design space and to reduce the number of code executions that are needed to find the optimum. We demonstrate constrained Bayesian optimization with a Pu-239/polyethylene solution system and a TEX experiment that is designed for critical ity safety validation of a nuclear waste model at the Hanford Site. For both systems, a global optimum was found within 75 Monte Carlo simulations.

42 ENGINEERING↗

Conceptual Design of a Scintillator-Based Fast-Ion Loss Detector for the Wendelstein 7-X Stellarator

Here, a conceptual design of a scintillator-based fast-ion loss detector (FILD) has been developed for the Wendelstein 7-X stellarator (W7-X). Simulations using the Monte Carlo codes ASCOT5 and BEAMS3D have been performed to calculate the expected flux of neutral beam injection (NBI)-generated fast hydrogen ions onto the conceptual detector probe head. These fast-ion loss fluxes have been calculated for several magnetic field configurations as well as probe insertion positions. At the maximum insertion position, both co-and counter-going losses with high incident pitch angles are observed; however, at retracted positions, only co-going fast ions reach the probe head. The FILDSIM code has been used to optimize the geometry of the detector entrance and collimating elements to achieve a wide velocity space coverage as well as a high velocity–space resolution. A synthetic FILD signal is calculated for the expected loss distribution via forward modeling using the instrument response function. The synthetic signal is found to largely retain the velocity space features of the loss distribution.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Multi-core and many-core shared-memory parallel raycasting volume rendering optimization and tuning

Given the computing industry trend of increasing processing capacity by adding more cores to a chip, the focus of this work is tuning the performance of a staple visualization algorithm, raycasting volume rendering, for shared-memory parallelism on multi-core CPUs and many-core GPUs. Our approach is to vary tunable algorithmic settings, along with known algorithmic optimizations and two different memory layouts, and measure performance in terms of absolute runtime and L2 memory cache misses. Our results indicate there is a wide variation in runtime performance on all platforms, as much as 254% for the tunable parameters we test on multi-core CPUs and 265% on many-core GPUs, and the optimal configurations vary across platforms, often in a non-obvious way. For example, our results indicate the optimal configurations on the GPU occur at a crossover point between those that maintain good cache utilization and those that saturate computational throughput. This result is likely to be extremely difficult to predict with an empirical performance model for this particular algorithm because it has an unstructured memory access pattern that varies locally for individual rays and globally for the selected viewpoint. Our results also show that optimal parameters on modern architectures are markedly different from those in previous studies run on older architectures. In addition, given the dramatic performance variation across platforms for both optimal algorithm settings and performance results, there is a clear benefit for production visualization and analysis codes to adopt a strategy for performance optimization through auto-tuning. These benefits will likely become more pronounced in the future as the number of cores per chip and the cost of moving data through the memory hierarchy both increase.

97 MATHEMATICS AND COMPUTING↗

Multi-core and Many-core Shared-memory Parallel Raycasting Volume Rendering Optimization and Tuning

Given the computing industry trend of increasing processing capacity by adding more cores to a chip, the focus of this work is tuning the performance of a staple visualization algorithm, raycasting volume rendering, for shared-memory parallelism on multi-core CPUs and many-core GPUs. Our approach is to vary tunable algorithmic settings, along with known algorithmic optimizations and two different memory layouts, and measure performance in terms of absolute runtime and L2 memory cache misses. Our results indicate there is a wide variation in runtime performance on all platforms, as much as 254% for the tunable parameters we test on multi-core CPUs and 265% on many-core GPUs, and the optimal configurations vary across platforms, often in a non-obvious way. For example, our results indicate the optimal configurations on the GPU occur at a crossover point between those that maintain good cache utilization and those that saturate computational throughput. This result is likely to be extremely difficult to predict with an empirical performance model for this particular algorithm because it has an unstructured memory access pattern that varies locally for individual rays and globally for the selected viewpoint. Our results also show that optimal parameters on modern architectures are markedly different from those in previous studies run on older architectures. And, given the dramatic performance variation across platforms for both optimal algorithm settings and performance results, there is a clear benefit for production visualization and analysis codes to adopt a strategy for performance optimization through auto-tuning. These benefits will likely become more pronounced in the future as the number of cores per chip and the cost of moving data through the memory hierarchy both increase.

97 MATHEMATICS AND COMPUTING↗