Search NASA⌕ Search

SEARCH · Search NASA

Results for “Optimal algorithms”

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 1,189 records · Page 66

A Transfer Learning Strategy for Improving the Data Efficiency of Deep Reinforcement Learning Control in Smart Buildings

Reinforcement learning (RL) is a powerful tool that has shown promising results in many domains such as robotics and game-playing. Because RL algorithms learn optimal control policies by continuously interacting with their environments, these algorithms require a lot of data to learn, which limits their application to a wide range of domains. For this reason, there is an immense need for improving the training and data efficiency of RL. Towards addressing this research gap, this paper proposes a transfer learning (TL) approach to improve the efficiency of the RL algorithms by reducing data need and, thus, reducing training time. To demonstrate the proposed approach, a knowledge transfer from a set of buildings to another building was conducted. The results show that the proposed TL approach is a promising method that can efficiently harness the information from similar RL tasks and reduce the data needs of RL algorithms.

Amasyali, Kadir↗

Precision Computations in Strongly Coupled Conformal Field Theories (Final Technical Report)

Conformal Field Theories (CFTs) are quantum field theories that are invariant under the conformal symmetry group (which includes translations and rotations, but also local rescalings of spacetime). They are building blocks of general quantum field theories, and appear in many areas of physics, including statistical physics, condensed matter physics, particle physics, and quantum gravity. Because of their extra symmetries, the mathematical structure of CFTs is tightly constrained, and this leads to the idea of the ``conformal bootstrap," which is to use these mathematical structures to constrain, and in some cases determine, CFT observables. A new numerical implementation of the conformal bootstrap idea appeared in 2008 with the work of Rattazzi, Rychkov, Tonni, and Vichi. Their observation was that certain bootstrap constraints (conformal symmetry and unitarity) could be combined to yield a convex optimization problem that constraints CFT data. By solving this convex optimization problem on a computer, one could obtain bounds on observables like critical exponents and operator product expansion (OPE) coefficients. Over the course of this award, the PI has improved numerical bootstrap techniques by optimizing known algorithms and finding new ones for performing the required convex optimization computations. The PI has applied these techniques to compute high-precision observables in several important strongly-coupled systems. The PI has also explored both analytical and numerical bootstrap methods for constraining the space of low energy effective field theories of quantum gravity, and developed new analytical techniques for CFT and QFT more broadly.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

MULTI-SCALE MODELING AND APPROXIMATION ASSISTED OPTIMIZATION OF BARE TUBE HEAT EXCHANGERS

Air-to-refrigerant heat exchangers are very common in air-conditioning, heat pump and refrigeration applications. In these heat exchangers, there is a great benefit in terms of size, weight, refrigerant charge and heat transfer coefficient, by moving from conventional channel sizes (~ 9mm) to smaller channel sizes (< 5mm). This work investigates new designs for air-to-refrigerant heat exchangers with tube outer diameter ranging from 0.5 to 2.0mm. The goal of this research is to develop and optimize the design of these heat exchangers and compare their performance with existing state of the art designs. The air-side performance of various tube bundle configurations are analyzed using a Parallel Parameterized CFD (PPCFD) technique. PPCFD allows for fast-parametric CFD analyses of various geometries with topology change. Approximation techniques drastically reduce the number of CFD evaluations required during optimization. Maximum Entropy Design method is used for sampling and Kriging method is used for metamodeling. Metamodels are developed for the air-side heat transfer coefficients and pressure drop as a function of tube-bundle dimensions and air velocity. The metamodels are then integrated with an air-to-refrigerant heat exchanger design code. This integration allows a multi-scale analysis of air-side performance heat exchangers including air-to-refrigerant heat transfer and phase change. Overall optimization is carried out using a multi-objective genetic algorithm. The optimal designs found can exhibit 50 percent size reduction, 75 percent decrease in air side pressure drop and doubled air heat transfer coefficients compared to a high performance compact micro channel heat exchanger with same capacity and flow rates.

CFD↗

Comparing implementations of penalized weighted least-squares sinogram restoration

Purpose: A CT scanner measures the energy that is deposited in each channel of a detector array by x rays that have been partially absorbed on their way through the object. The measurement process is complex and quantitative measurements are always and inevitably associated with errors, so CT data must be preprocessed prior to reconstruction. In recent years, the authors have formulated CT sinogram preprocessing as a statistical restoration problem in which the goal is to obtain the best estimate of the line integrals needed for reconstruction from the set of noisy, degraded measurements. The authors have explored both penalized Poisson likelihood (PL) and penalized weighted least-squares (PWLS) objective functions. At low doses, the authors found that the PL approach outperforms PWLS in terms of resolution-noise tradeoffs, but at standard doses they perform similarly. The PWLS objective function, being quadratic, is more amenable to computational acceleration than the PL objective. In this work, the authors develop and compare two different methods for implementing PWLS sinogram restoration with the hope of improving computational performance relative to PL in the standard-dose regime. Sinogram restoration is still significant in the standard-dose regime since it can still outperform standard approaches and it allows for correction of effects that are not usually modeled in standard CT preprocessing. Methods: The authors have explored and compared two implementation strategies for PWLS sinogram restoration: (1) A direct matrix-inversion strategy based on the closed-form solution to the PWLS optimization problem and (2) an iterative approach based on the conjugate-gradient algorithm. Obtaining optimal performance from each strategy required modifying the naive off-the-shelf implementations of the algorithms to exploit the particular symmetry and sparseness of the sinogram-restoration problem. For the closed-form approach, the authors subdivided the large matrix inversion into smaller coupled problems and exploited sparseness to minimize matrix operations. For the conjugate-gradient approach, the authors exploited sparseness and preconditioned the problem to speed up convergence. Results: All methods produced qualitatively and quantitatively similar images as measured by resolution-variance tradeoffs and difference images. Despite the acceleration strategies, the direct matrix-inversion approach was found to be uncompetitive with iterative approaches, with a computational burden higher by an order of magnitude or more. The iterative conjugate-gradient approach, however, does appear promising, with computation times half that of the authors' previous penalized-likelihood implementation. Conclusions: Iterative conjugate-gradient based PWLS sinogram restoration with careful matrix optimizations has computational advantages over direct matrix PWLS inversion and over penalized-likelihood sinogram restoration and can be considered a good alternative in standard-dose regimes.

60 APPLIED LIFE SCIENCES↗

Algorithm 1028: VTMOP: Solver for Blackbox Multiobjective Optimization Problems

VTMOP is a Fortran 2008 software package containing two Fortran modules for solving computationally expensive bound-constrained blackbox multiobjective optimization problems. VTMOP implements the algorithm of [32], which handles two or more objectives, does not require any derivatives, and produces well-distributed points over the Pareto front. The first module contains a general framework for solving multiobjective optimization problems by combining response surface methodology, trust region methodology, and an adaptive weighting scheme. The second module features a driver subroutine that implements this framework when the objective functions can be wrapped as a Fortran subroutine. Lastly, support is provided for both serial and parallel execution paradigms, and VTMOP is demonstrated on several test problems as well as one real-world problem in the area of particle accelerator optimization.

97 MATHEMATICS AND COMPUTING↗

Optimizing the Accelerated Recursive Doubling Algorithm for Block Tridiagonal Systems of Equations

The need to solve block tridiagonal systems with hundreds or thousands of right-hand sides for the same block tridiagonal matrix is common in a variety of disciplines. To meet this need, the Accelerated Recursive Doubling Algorithm was developed. After a right-hand side independent phase, the algorithm allows for the quick, online calculation of solutions for different right-hand sides. In this work, we present methods to optimize the Accelerated Recursive Doubling Algorithm in memory usage and computation time in a hybrid parallelization model. The right-hand side independent phase of the naïve implementation takes ≥ 11/3 the amount of memory required to store the tridiagonal matrix, while our implementation reduces the fraction to ≈ 5/3 . The right-hand side dependent phase of the naïve implementation takes ≥ 6 times the amount of memory required to store the right-hand side, while our implementation reduces the fraction to ≈ 3. The computation time for the independent phase is reduced to ≈ 2/3 times that of the naïve implementation, while the computation time for the dependent phase is reduced to ≈ 5/9 . With increasing numbers of shared-memory threads q on every distributed processing element, we have O(q) theoretical speedup.

97 MATHEMATICS AND COMPUTING↗

Gradient-Informed Design Optimization of Select Nuclear Systems

In this work, we present a gradient-informed design optimization of nuclear reactor core components based on neutronics objectives with both continuous and discrete materials. The main argument in favor of using gradient-informed design optimization is that it scales well with increasing dimensionality of the design space. First, a challenge problem with 121 free parameters is solved with a gradient-informed method and then with a genetic algorithm. Then, a challenge problem to optimize the flux profile of a simplified assembly with eight axial zones is solved. Both challenge problems are solved using directly calculated derivatives from Tools for Sensitivity and Uncertainty Analysis Methodology Implementation (TSUNAMI) in the SCALE package. Furthermore, we demonstrate how a discrete optimization problem—selection of materials for 121 voxels—can be lifted into a continuous problem with mixed materials. In the continuous space, adjoint-based gradients are well-defined, and gradient descent is applicable. Then, a forcing function is introduced that with the selection of an appropriately sized hyperparameter can be used to guide the optimized continuous solution back into a discrete solution. This paper presents an account of the challenges that were faced when applying a gradient-informed optimization algorithm using a Monte Carlo calculation to estimate the gradient information and compares a gradient descent optimization method to a genetic algorithm optimization of the same geometry. Overall, this work demonstrates the potential use of adjoint-based gradient calculations in design optimization of nuclear systems.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗

Gains in operational flexibility, safety margins, and cost efficiencies via integrated Plant Reload Optimization platform

The U.S. Department of Energy Light Water Reactor Sustainability Program Risk-Informed Systems Analysis Pathway Plant Reload Optimization Project aims to develop an integrated, comprehensive framework offering an all-in-one solution for reload evaluations with a special focus on optimizing core design. Optimizing the fuel loading pattern is one of the most important considerations in reducing the amount of new fuel used in the core. Due to thousands of possible core configuration options, finding optimal solutions is an unachievable task for a human. The Plant ReLoad Optimization platform, which supports artificial-intelligence-based reactor core designing, is now fully capable of handling realistic problems. The Plant ReLoad Optimization platform development project aims to build a reactor core design tool that includes reactor safety and fuel performance analyses and uses artificial intelligence to support the optimization of core design solutions. The NSGA-II (Non-dominated Sorting Genetic Algorithm II) optimizer was developed and tested within RAVEN (Risk Analysis and Virtual ENvironment) to handle many constraints by using an augmented objectives methodology. The demonstration was performed with constrained multiobjective optimization of a 17 × 17 pressurized-water reactor core loading patterns to minimize fuel cost and maximize fuel cycle length.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗

Quantum dot ternary-valued full-adder: Logic synthesis by a multiobjective design optimization based on a genetic algorithm

A methodology is proposed for designing a low-energy consuming ternary-valued full adder based on a quantum dot (QD) electrostatically coupled with a single electron transistor operating as a charge sensor. The methodology is based on design optimization: the values of the physical parameters of the system required for implementing the logic operations are optimized using a multiobjective genetic algorithm. The searching space is determined by elements of the capacitance matrix describing the electrostatic couplings in the entire device. The objective functions are defined as the maximal absolute error over actual device logic outputs relative to the ideal truth tables for the sum and the carry-out in base 3. The logic units are implemented on the same device: a single dual-gate quantum dot and a charge sensor. Their physical parameters are optimized to compute either the sum or the carry out outputs and are compatible with current experimental capabilities. The outputs are encoded in the value of the electric current passing through the charge sensor, while the logic inputs are supplied by the voltage levels on the two gate electrodes attached to the QD. The complex logic ternary operations are directly implemented on an extremely simple device, characterized by small sizes and low-energy consumption compared to devices based on switching single-electron transistors. The design methodology is general and provides a rational approach for realizing non-switching logic operations on QD devices.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

WE-AB-209-06: Dynamic Collimator Trajectory Algorithm for Use in VMAT Treatment Deliveries

Purpose: To develop advanced dynamic collimator positioning algorithms for optimal beam’s-eye-view (BEV) fitting of targets in VMAT procedures, including multiple metastases stereotactic radiosurgery procedures. Methods: A trajectory algorithm was developed, which can dynamically modify the angle of the collimator as a function of VMAT control point to provide optimized collimation of target volume(s). Central to this algorithm is a concept denoted “whitespace”, defined as area within the jaw-defined BEV field, outside of the PTV, and not shielded by the MLC when fit to the PTV. Calculating whitespace at all collimator angles and every control point, a two-dimensional topographical map depicting the tightness-of-fit of the MLC was generated. A variety of novel searching algorithms identified a number of candidate trajectories of continuous collimator motion. Ranking these candidate trajectories according to their accrued whitespace value produced an optimal solution for navigation of this map. Results: All trajectories were normalized to minimum possible (i.e. calculated without consideration of collimator motion constraints) accrued whitespace. On an acoustic neuroma case, a random walk algorithm generated a trajectory with 151% whitespace; random walk including a mandatory anchor point improved this to 148%; gradient search produced a trajectory with 137%; and bi-directional gradient search generated a trajectory with 130% whitespace. For comparison, a fixed collimator angle of 30° and 330° accumulated 272% and 228% of whitespace, respectively. The algorithm was tested on a clinical case with two metastases (single isocentre) and identified collimator angles that allow for simultaneous irradiation of the PTVs while minimizing normal tissue irradiation. Conclusion: Dynamic collimator trajectories have the potential to improve VMAT deliveries through increased efficiency and reduced normal tissue dose, especially in treatment of multiple cranial metastases, without significant safety concerns that hinder immediate clinical implementation.

60 APPLIED LIFE SCIENCES↗

Extremum seeking-based optimization of high voltage converter modulator rise-time

We digitally implement an extremum seeking (ES) algorithm, which optimizes the rise time of the output voltage of a high voltage converter modulator (HVCM) at the Los Alamos Neutron Science Center (LANSCE) HVCM test stand by iteratively, simultaneously tuning the first 8 switching edges of each of the three phase drive waveforms (24 variables total). We achieve a 50 μs rise time, which is reduction in half compared to the 100 μs achieved at the Spallation Neutron Source (SNS) at Oak Ridge National Laboratory. Considering that HVCMs typically operate with an output voltage of 100 kV, with a 60Hz repetition rate, the 50 μs rise time reduction will result in very significant energy savings. The ES algorithm will prove successful, despite the noisy measurements and cost calculations, confirming the theoretical results that the algorithm is not affected by noise whose frequency components are independent of the perturbing frequencies.

43 PARTICLE ACCELERATORS↗

Characterization of BNL and HPK AC-LGAD sensors with a 120 GeV proton beam

We present measurements of AC-LGADs performed at the Fermilab's test beam facility using 120 GeV protons. Here we studied the performance of various strip and pad AC-LGAD sensors that were produced by BNL and HPK. The measurements are performed with our upgraded test beam setup that utilizes a high precision telescope tracker, and a simultaneous readout of up to 7 channels per sensor, which allows detailed studies of signal sharing characteristics. These measurements allow us to assess the differences in designs between different manufacturers, and optimize them based on experimental performance. We then study several reconstruction algorithms to optimize position and time resolutions that utilize the signal sharing properties of each sensor. We present a world's first demonstration of silicon sensors in a test beam that simultaneously achieve better than 6–10 μm position and 30 ps time resolution. This represents a substantial improvement to the spatial resolution than would be obtained with binary readout of sensors with similar pitch.

47 OTHER INSTRUMENTATION↗

Protection settings optimizer

SAND2023-06672O The Protection Settings Optimizer (PSO) uses system and fault data as inputs to formulate the problem of calculating relay settings as a mixed integer, nonlinear optimization problem (MINLP). The MINLP is solved using a genetic algorithm-based optimizer that attempts to find settings to reduce the relay operating times. The PSO protects the power system by using the steady-state fault voltages and currents, which then calculates the optimal device setting to protect the power system. Sandia National Laboratories is a multimission laboratory managed and operated by National Technology & Engineering Solutions of Sandia, LLC, a wholly owned subsidiary of Honeywell International Inc., for the U.S. Department of Energy’s National Nuclear Security Administration under contract DE-NA0003525.

Patel, Trupal↗

Characterization of BNL and HPK AC-LGAD sensors with a 120 GeV proton beam

We present measurements of AC-LGADs performed at the Fermilab's test beam facility using 120 GeV protons. We studied the performance of various strip and pad AC-LGAD sensors that were produced by BNL and HPK. The measurements are performed with our upgraded test beam setup that utilizes a high precision telescope tracker, and a simultaneous readout of up to 7 channels per sensor, which allows detailed studies of signal sharing characteristics. These measurements allow us to assess the differences in designs between different manufacturers, and optimize them based on experimental performance. We then study several reconstruction algorithms to optimize position and time resolutions that utilize the signal sharing properties of each sensor. We present a world's first demonstration of silicon sensors in a test beam that simultaneously achieve better than 6-10 micron position and 30 ps time resolution. This represents a substantial improvement to the spatial resolution than would be obtained with binary readout of sensors with similar pitch.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Proximal Galerkin: A Structure-Preserving Finite Element Method for Pointwise Bound Constraints

The proximal Galerkin finite element method is a high-order, low iteration complexity, nonlinear numerical method that preserves the geometric and algebraic structure of pointwise bound constraints in infinite-dimensional function spaces. This paper introduces the proximal Galerkin method and applies it to solve free boundary problems, enforce discrete maximum principles, and develop a scalable, mesh-independent algorithm for optimal design with pointwise bound constraints. This paper also introduces the latent variable proximal point (LVPP) algorithm, from which the proximal Galerkin method derives. When analyzing the classical obstacle problem, we discover that the underlying variational inequality can be replaced by a sequence of second-order partial differential equations (PDEs) that are readily discretized and solved with, e.g., the proximal Galerkin method. Throughout this work, we arrive at several contributions that may be of independent interest. These include (1) a semilinear PDE we refer to as the entropic Poisson equation; (2) an algebraic/geometric connection between high-order positivity-preserving discretizations and certain infinite-dimensional Lie groups; and (3) a gradient-based, bound-preserving algorithm for two-field, density-based topology optimization. The complete proximal Galerkin methodology combines ideas from nonlinear programming, functional analysis, tropical algebra, and differential geometry and can potentially lead to new synergies among these areas as well as within variational and numerical analysis. Open-source implementations of our methods accompany this work to facilitate reproduction and broader adoption.

97 MATHEMATICS AND COMPUTING↗

Optimal sensor network design for multi-scale, time-varying differential algebraic equation systems: Application to an entrained-flow gasifier refractory brick

An algorithm for optimal sensor network design for multi-scale, time-varying differential algebraic equation systems with non-separable dynamics is presented. As the process is time-varying, an integral normalized posterior error covariance of a multi-scale filter is minimized to obtain the optimal sensor locations. For reducing the computational cost, an adaptive sampling rate approach is considered for the slowly-varying variables. The algorithm is applied to a smart refractory brick with embedded sensors as part of an entrained-flow gasifier. Thermistors and interdigital capacitors are considered as candidate measurement technologies for estimating temperature and slag penetration profile along the gasifier wall. Finally, when the optimal set of sensors obtained from the algorithm is used for estimating temperature and slag penetration profiles in a multi-scale Kalman filter framework, satisfactory estimates are obtained despite high measurement noise and model mismatch.

42 ENGINEERING↗

OpenMP Parallelization and Optimization of Graph-Based Machine Learning Algorithms

In this paper, we investigate the OpenMP parallelization and optimization of two novel data classification algorithms. The new algorithms are based on graph and PDE solution techniques and provide significant accuracy and performance advantages over traditional data classification algorithms in serial mode. The methods leverage the Nystrom extension to calculate eigenvalue/eigenvectors of the graph Laplacian and this is a self-contained module that can be used in conjunction with other graph-Laplacian based methods such as spectral clustering. We use performance tools to collect the hotspots and memory access of the serial codes and use OpenMP as the parallelization language to parallelize the most time-consuming parts. Where possible, we also use library routines. We then optimize the OpenMP implementations and detail the performance on traditional supercomputer nodes (in our case a Cray XC30), and test the optimization steps on emerging testbed systems based on Intel’s Knights Corner and Landing processors. We show both performance improvement and strong scaling behavior. Finally, a large number of optimization techniques and analyses are necessary before the algorithm reaches almost ideal scaling.

97 MATHEMATICS AND COMPUTING↗