Search NASA⌕ Search

SEARCH · Search NASA

Results for “optimization algorithm”

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 217 records · Page 12
All sourcesNASA NTRSDOE OSTI
DOE OSTI2013

A Case for Soft Error Detection and Correction in Computational Chemistry

High performance computing platforms are expected to deliver 10(18) floating operations per second by the year 2022 through the deployment of millions of cores. Even if every core is highly reliable the sheer number of the them will mean that the mean time between failures will become so short that most applications runs will suffer at least one fault. In particular soft errors caused by intermittent incorrect behavior of the hardware are a concern as they lead to silent data corruption. In this paper we investigate the impact of soft errors on optimization algorithms using Hartree-Fock as a particular example. Optimization algorithms iteratively reduce the error in the initial guess to reach the intended solution. Therefore they may intuitively appear to be resilient to soft errors. Our results show that this is true for soft errors of small magnitudes but not for large errors. We suggest error detection and correction mechanisms for different classes of data structures. The results obtained with these mechanisms indicate that we can correct more than 95% of the soft errors at moderate increases in the computational cost.

Computational chemistry↗
DOE OSTI2014

Interior Point Algorithms for Optimization Problems (Final Technical Report)

Over the period of this award we developed an algorithmic framework for constraint reduction in linear programming (LP) and convex quadratic programming (QP), proved convergence of our algorithms, and applied them to a variety of applications, including entropy-based moment closure in gas dynamics.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2019

Near-Optimal and Practical Algorithms for Graph Scan Statistics with Connectivity Constraints

One fundamental task in network analysis is detecting “hotspots” or “anomalies” in the network; that is, detecting subgraphs where there is significantly more activity than one would expect given historical data or some baseline process. Scan statistics is one popular approach used for anomalous subgraph detection. This methodology involves maximizing a score function over all connected subgraphs, which is a challenging computational problem. A number of heuristics have been proposed for these problems, but they do not provide any quality guarantees. Here, we propose a framework for designing algorithms for optimizing a large class of scan statistics for networks, subject to connectivity constraints. Our algorithms run in time that scales linearly on the size of the graph and depends on a parameter we call the “effective solution size,” while providing rigorous approximation guarantees. In contrast, most prior methods have super-linear running times in terms of graph size. Extensive empirical evidence demonstrates the effectiveness and efficiency of our proposed algorithms in comparison with state-of-the-art methods. Our approach improves on the performance relative to all prior methods, giving up to over 25% increase in the score. Further, our algorithms scale to networks with up to a million nodes, which is 1–2 orders of magnitude larger than all prior applications.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2014

Optimized Uncertainty Quantification Algorithm Within a Dynamic Event Tree Framework

Methods for developing Phenomenological Identification and Ranking Tables (PIRT) for nuclear power plants have been a useful tool in providing insight into modelling aspects that are important to safety. These methods have involved expert knowledge with regards to reactor plant transients and thermal-hydraulic codes to identify are of highest importance. Quantified PIRT provides for rigorous method for quantifying the phenomena that can have the greatest impact. The transients that are evaluated and the timing of those events are typically developed in collaboration with the Probabilistic Risk Analysis. Though quite effective in evaluating risk, traditional PRA methods lack the capability to evaluate complex dynamic systems where end states may vary as a function of transition time from physical state to physical state . Dynamic PRA (DPRA) methods provide a more rigorous analysis of complex dynamic systems. A limitation of DPRA is its potential for state or combinatorial explosion that grows as a function of the number of components; as well as, the sampling of transition times from state-to-state of the entire system. This paper presents a method for performing QPIRT within a dynamic event tree framework such that timing events which result in the highest probabilities of failure are captured and a QPIRT is performed simultaneously while performing a discrete dynamic event tree evaluation. The resulting simulation results in a formal QPIRT for each end state. The use of dynamic event trees results in state explosion as the number of possible component states increases. This paper utilizes a branch and bound algorithm to optimize the solution of the dynamic event trees. The paper summarizes the methods used to implement the branch-and-bound algorithm in solving the discrete dynamic event trees.

42 ENGINEERING↗
DOE OSTI2025

Randomized Adiabatic Quantum Linear Solver Algorithm with Optimal Complexity Scaling and Detailed Running Costs

Solving linear systems of equations is a fundamental problem with a wide variety of applications across many fields of science, and there is increasing effort to develop quantum linear solver algorithms. Subaşı et al. [Phys. Rev. Lett. 122, 060504 (2019)] proposed a randomized algorithm inspired by adiabatic quantum computing, based on a sequence of random Hamiltonian simulation steps, with suboptimal scaling in the condition number 𝜅 of the linear system and the target error 𝜖. Here we go beyond these results in several ways. Firstly, using filtering [Lin and Tong, Quantum 4, 361 (2020)] and Poissonization techniques [Cunningham and Roland, ArXiv:2406.03972 (2024)], the algorithm complexity is improved to the optimal scaling 𝑂⁡(𝜅⁢log (1/𝜖))—an exponential improvement in 𝜖, and a shaving of a log 𝜅 scaling factor in 𝜅. Secondly, the algorithm is further modified to achieve constant factor improvements, which are vital as we progress towards hardware implementations on fault-tolerant devices. We introduce a cheaper randomized walk operator method replacing Hamiltonian simulation—which also removes the need for potentially challenging classical precomputations; randomized routines are sampled over optimized random variables; circuit constructions are improved. We obtain a closed formula rigorously upper bounding the expected number of times one needs to apply a block-encoding of the linear system matrix to output a quantum state encoding the solution to the linear system. The upper bound is 837⁢𝜅 at 𝜖 = 10 −10 for Hermitian matrices.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2013

PERFORMANCE OF A NOVEL FAST TRANSIENTS DETECTION SYSTEM

We investigate the signal-to-noise ratio (S/N) of a new incoherent dedispersion algorithm optimized for FPGA-based architectures intended for deployment on the Australian SKA Pathfinder and other Square Kilometre Array precursors for fast transients surveys. Unlike conventional CPU- and GPU-optimized incoherent dedispersion algorithms, this algorithm has the freedom to maximize the S/N by way of programmable dispersion profiles that enable the inclusion of different numbers of time samples per spectral channel. This allows, for example, more samples to be summed at lower frequencies where intra-channel dispersion smearing is larger, or it could even be used to optimize the dedispersion sum for steep spectrum sources. Our analysis takes into account the intrinsic pulse width, scatter broadening, spectral index and dispersion measure of the signal, and the system's frequency range, spectral and temporal resolution, and number of trial dedispersions. We show that the system achieves better than 80% of the optimal S/N where the temporal resolution and the intra-channel smearing time are smaller than a quarter of the average width of the pulse across the system's frequency band (after including scatter smearing). Coarse temporal resolutions suffer a {Delta}t {sup -1/2} decay in S/N, and coarse spectral resolutions cause a {Delta}{nu}{sup -1/2} decay in S/N, where {Delta}t and {Delta}{nu} are the temporal and spectral resolutions of the system, respectively. We show how the system's S/N compares with that of matched filter and boxcar filter detectors. We further present a new algorithm for selecting trial dispersion measures for a survey that maintains a given minimum S/N performance across a range of dispersion measures.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2023

Structure optimization with stochastic density functional theory

Linear-scaling techniques for Kohn–Sham density functional theory are essential to describe the ground state properties of extended systems. Still, these techniques often rely on the localization of the density matrix or accurate embedding approaches, limiting their applicability. In contrast, stochastic density functional theory (sDFT) achieves linear- and sub-linear scaling by statistically sampling the ground state density without relying on embedding or imposing localization. In return, ground state observables, such as the forces on the nuclei, fluctuate in sDFT, making optimizing the nuclear structure a highly non-trivial problem. In this work, we combine the most recent noise-reduction schemes for sDFT with stochastic optimization algorithms to perform structure optimization within sDFT. We compare the performance of the stochastic gradient descent approach and its variations (stochastic gradient descent with momentum) with stochastic optimization techniques that rely on the Hessian, such as the stochastic Broyden–Fletcher–Goldfarb–Shanno algorithm. In conclusion, we further provide a detailed assessment of the computational efficiency and its dependence on the optimization parameters of each method for determining the ground state structure of bulk silicon with varying supercell dimensions.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2022

Development of novel dynamic machine learning-based optimization of a coal-fired power plant

The increasing fraction of intermittent renewable energy in the electrical grid is resulting in coal-fired boilers now routinely ramp up and down. The current state-of-the-art operation for such boilers is to apply steady-state, neural network-based optimization to make control decisions in real-time, and this report demonstrates the feasibility of extending this to dynamic, neural network-based optimization using a long short-term memory neural network. A simplified numerical simulation of a t-fired coal boiler and supporting equipment is used to represent a real plant subjected to both steady-state, neural network-based optimization and dynamic, neural network-based optimization. Using the same intervals and a particle swarm optimization algorithm, the dynamic optimization outperforms the steady-state optimization and realizes up to 4.58% improvement in thermal efficiency. Dynamic optimization with a long short-term memory neural network is shown to both be feasible and beneficial for operation of a coal-fired boiler under changing load.

42 ENGINEERING↗
DOE OSTI2024

Optimizing and Exploring Untapped Micro-Hydro Hybrid Systems: a Multi-Objective Approach for Crystal Lake as a Large-Scale Energy Storage Solution

Increasing electricity demand and concerns about climate change and fossil fuel consumption have highlighted the importance of renewable energy resources and storage systems. This paper proposes a method for exploring untapped pumped hydro storage potentials to accommodate intermittent renewable energy generation profiles. Hourly measured data from 2022 in Benzie County, Michigan, United States, were gathered for system sizing and a thorough, realistic analysis. By employing the multi-objective grey wolf optimization algorithm, we formulated optimal sizing and energy-management strategies for three different scenarios. Unlike similar studies, the 3rd with triple objective functions (OFs) scenario aims to maximize both reliability and ecological OFs while minimizing the cost OF. It has shown promising results with multiple solutions, considering economic, environmental, and reliability factors. A case study conducted in Crystal Lake, Michigan, revealed that although Crystal Lake would function only as a micro-hydro power facility, it is a promising and huge storage unit with a substantial storage capacity of around 14.9734GWh. The system investigated is significant in the USA due to its rapid deployment capabilities, minimal construction requirements, and ease of integration with the distribution grid. The fuzzy logic method was employed to identify the best non-dominant solution among the other solutions. Furthermore, these outcomes include a notably low levelized cost of energy at 0.046147$/kWh, a robust index of reliability of 99.705%, and a significant reduction in CO₂ emissions amounting to 7.9142×10 3 tons/year, when considering the triple OFs. The paper’s methodology provides valuable insights for regions aiming to utilize renewable energy from untapped storage sources.

13 HYDRO ENERGY↗
DOE OSTI2021

Quantifying particle-scale 3D granular dynamics during rapid compaction from time-resolved in situ 2D x-ray images

Understanding the particle-scale dynamics of granular materials during rapid compaction and flow is of fundamental importance for manufacturing, planetary science, geology, and defense applications. Time-resolved 2D radiography and static 3D x-ray tomography are powerful in situ tools for studying particle-scale dynamics but provide detail only in 2D or with significant time-scale limitations, respectively. Here, we introduce a new method that uses 2D in situ x-ray imaging for determining time-resolved 3D particle-scale dynamics in rapidly compressed granular materials. The method employs initial particle packing structures obtained from x-ray tomography, a 2D x-ray image generation algorithm, and an optimization algorithm. We first describe and validate the method using finite element simulations. Furthermore, we then apply the technique to x-ray phase-contrast images obtained during rapid compaction of granular materials with varying particle morphology and sample thickness. The depth-resolved particle-scale dynamics reveal complex velocity and porosity fields evolving heterogeneously along and perpendicular to the compaction direction. We characterize these features, their fluctuations near the compaction front, and the compaction front thickness. Our technique can be applied to understanding granular dynamics during rapid compaction events, and rearrangements during slower, but non-quasi-static, flows.

36 MATERIALS SCIENCE↗
DOE OSTI2021

Cost Drivers for a Tokamak-Based Compact Pilot Plant

A comprehensive systems code that includes a range of physics and engineering considerations along with a simplified costing model has been utilized to evaluate the primary cost drivers for a compact tokamak pilot plant. The systems code has been benchmarked against several tokamak reactor designs and is utilized with sophisticated optimization algorithms to develop optimal solutions for a set of user-specified assumptions and design constraints. Here, in contrast to previous models that have focused on the cost of electricity as the key cost metric, this study uses the estimated capital cost of the facility. The analysis suggests that a pilot plant with the following features may offer potential for a cost-attractive pilot plant: A ~ 3, H 98y2 > 1.5, P net = 200 MW, 𝜏 𝑝⁢𝑢⁢𝑙⁢𝑠⁢𝑒 ~ 1 to 2 h utilizing rare-earth barium copper oxide (REBCO) magnet technology and a plug-bucked central solenoid/toroidal field (CS/TF) magnet support structure. While REBCO magnets offer some advantages relative to Nb 3 Sn magnets in all cases, the most gain is obtained when combined with the plug-bucked CS/TF bucking solution. Pulsed operation reduces capital cost requirements relative to steady-state operation, especially at low confinement. Cost sensitivity studies indicate that there are significant cost uncertainties associated with the achievable confinement quality, tritium breeding capability, attainable thermal efficiency, and achievable neutron wall loading, suggesting that these areas are the most critical areas in reducing the cost risk for a compact tokamak pilot plant. Further cost sensitivity studies indicate that the estimated cost is most sensitive to the underlying cost of the magnetic coils, providing further impetus to better establish cost-effective means for producing fusion magnets.

Compact pilot plant↗
DOE OSTI2022

Reaction Rover

Reaction Rover is a software application designed to enable reaction optimization with little or no human intervention and to provide a flexible platform to facilitate rapid testing, automation, and implementation of new reaction optimization algorithms. Toward these goals, Reaction Rover is designed to be modular, allowing the optimization algorithm, analyzer, and reactor components to be changed for any given process using easily integrated plugins. The Reaction Rover platform is a key development that enables the testing, use, and automation of machine learning (ML) and non-ML optimizers in a variety of laboratory situations.

Windus, Theresa↗
DOE OSTI2021

Enhanced Integer Permutation based Genetic Algorithm for Optimization of Tube-Fin Heat Exchanger Circuitry with Splits and Merges

Tube-fin heat exchangers (HXs) are widely used in air-conditioning and heat pump applications. The performance of these heat exchangers is strongly influenced by the refrigerant circuitry. Studies have proved that by optimizing the refrigerant circuitry, the performance of HXs can be significantly improved. In our previous research, an Integer Permutation based Genetic Algorithm (IPGA) was developed to obtain the optimal circuitry designs. Our previous research showed that IPGA demonstrates superior capability to obtain better refrigerant circuitries with lower computational cost than the other methods in literature. And the optimal circuitry designs obtained from IPGA are manufacturable with the available tooling. However, the IPGA developed previously cannot generate designs with splitting and merging of circuits. To remedy this limitation, a new chromosome which can represent circuitry with splitting and merging of circuits is developed. In addition to the six genetic operators implemented previously, two new genetic operators are developed to generate splits and merges. As a result, the enhanced IPGA can explore the solution space more thoroughly than the previous IPGA. A case study using an evaporator from an A-type indoor unit shows that, given the similar capacity improvements obtained from the enhanced IPGA compared with the previous IPGA, the refrigerant pressure drop reduction obtained from the enhanced IPGA is 26.5% compared against 1.0% pressure drop reduction from the previous IPGA. The benchmark of the enhanced IPGA with other methods in literature demonstrates that the enhanced IPGA can generate circuitry designs with performance superior to those obtained from other methods.

LI, Zhenning↗
DOE OSTI2015

A junction-tree based learning algorithm to optimize network wide traffic control: A coordinated multi-agent framework

Our study develops a novel reinforcement learning algorithm for the challenging coordinated signal control problem. Traffic signals are modeled as intelligent agents interacting with the stochastic traffic environment. The model is built on the framework of coordinated reinforcement learning. The Junction Tree Algorithm (JTA) based reinforcement learning is proposed to obtain an exact inference of the best joint actions for all the coordinated intersections. Moreover, the algorithm is implemented and tested with a network containing 18 signalized intersections in VISSIM. Finally, our results show that the JTA based algorithm outperforms independent learning (Q-learning), real-time adaptive learning, and fixed timing plans in terms of average delay, number of stops, and vehicular emissions at the network level.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2019

Performance of biologically inspired algorithms tuned on TiO 2 nanoparticle benchmark system

Biologically inspired optimization algorithms (BIOA), such as particle swarm optimization (PSO) and differential evolution (DE), are often applied to a specific problem with the tacit assumptions that each algorithm will have a unique set of “ideal” tuning parameters that provide “best” performance, and that, once tuned, two different optimization algorithms ought to exhibit distinct performance. This work tests those assumptions by systematically comparing PSO and DE applied to identify stable and metastable configurations of a benchmark system, a titania dimer Ti 2 O 4 , whose energy is described by a modified Matsui-Akaogi pairwise potential. The hypotheses (H) are formally expressed as: (H1) for both PSO and DE there exists a unique, optimal set of tuning parameters that maximize its performance on a given problem; (H2) when running two different BIOAs, one algorithm must demonstrate superior performance over the other. By identifying and studying an approximate Pareto-optimal set of tuning parameters for DE and PSO, H1 is shown to be strictly false for DE, and tentatively false for PSO, while H2 is shown to be false with respect to certain statistics. Here, we further investigate the correlations between key tuning parameters and algorithm performance. Our conclusion may be implemented to further optimize and develop BIOAs applied to materials systems.

36 MATERIALS SCIENCE↗
DOE OSTI2024

Multiple aspects maintenance ontology-based intelligent maintenance optimization framework for safety-critical systems

Abstract Maintenance optimization is a process for improving the efficiency of maintenance strategies and activities, considering various aspects of the target system and components, such as the probabilities of system failures and the cost of repair and replacement of a failed component. The improvement of maintenance optimization algorithms generally requires information from various data sources. For example, it may require the system risk information derived from risk analysis tools or the residual lifetime of a component from fault prognosis tools. The requirements of data acquisition (DAQ) and aggregation pose new challenges for maintenance management systems (MMSs) that implement and use these maintenance optimization algorithms. This paper proposes a multiple aspects maintenance ontology-based framework to facilitate DAQ from MMSs, online monitoring systems, fault detection and discrimination tools, risk assessment tools, decision-making tools, and component identification tools, and accelerate the implementation and verification of contemporary maintenance optimization models and algorithms. The proposed framework consists of a multi-aspect maintenance ontology with critical information for maintenance optimization and application interfaces for collecting information from various data sources, such as fault prognosis tools, online monitoring tools, risk assessment tools, and decision-making algorithms. In addition, this paper proposes a heuristic method for integrating concepts and properties from other existing ontologies into the proposed framework when the existing ontology is not fully compatible with the ontology under construction. Finally, the paper verifies the proposed ontology framework using a feedwater system designed for nuclear power plants with valves and filters as the components under maintenance.

Diao, Xiaoxu (ORCID:0000000346726352)↗