Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Global optimization”

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

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

At least 1,063 records · Page 59
All sourcesNASA NTRSDOE OSTI
NASA NTRS1984

On the utilization of engineering knowledge in design optimization

Some current research work conducted at the University of Michigan is described to illustrate efforts for incorporating knowledge in optimization in a nontraditional way. The incorporation of available knowledge in a logic structure is examined in two circumstances. The first examines the possibility of introducing global design information in a local active set strategy implemented during the iterations of projection-type algorithms for nonlinearly constrained problems. The technique used algorithms for nonlinearly constrained problems. The technique used combines global and local monotinicity analysis of the objective and constraint functions. The second examines a knowledge-based program which aids the user to create condigurations that are most desirable from the manufacturing assembly viewpoint. The data bank used is the classification scheme suggested by Boothroyd. The important aspect of this program is that it is an aid for synthesis intended for use in the design concept phase in a way similar to the so-called idea-triggers in creativity-enhancement techniques like brain-storming. The idea generation, however, is not random but it is driven by the goal of achieving the best acceptable configuration.

Papalambros, P.↗
DOE OSTI2023

A Sequential Quadratic Programming Algorithm for Nonsmooth Problems with Upper- \({\boldsymbol{\mathcal{C}^2}}\) Objective

An optimization algorithm for nonsmooth nonconvex constrained optimization problems with upper- \({\boldsymbol{\mathcal{C}^2}}\) objective functions is proposed and analyzed. Upper- \({\boldsymbol{\mathcal{C}^2}}\) is a weakly concave property that exists in difference of convex (DC) functions and arises naturally in many applications, particularly certain classes of solutions to parametric optimization problems e.g., recourse of stochastic programming and projection onto closed sets. The algorithm can be viewed as an extension of sequential quadratic programming (SQP) to nonsmooth problems with upper- \({\boldsymbol{\mathcal{C}^2}}\) objectives or a simplified bundle method. It is globally convergent with bounded algorithm parameters that are updated with a trust-region criterion. The algorithm handles general smooth constraints through linearization and uses a line search to ensure progress. The potential inconsistencies from the linearization of the constraints are addressed through a penalty method. In conclusion, the capabilities of the algorithm are demonstrated by solving both simple upper- \({\boldsymbol{\mathcal{C}^2}}\) problems and a real-world optimal power flow problem used in current power grid industry practices.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2023

An Incremental Gradient Method for Optimization Problems With Variational Inequality Constraints

We consider minimizing a sum of agent-specific nondifferentiable merely convex functions over the solution set of a variational inequality (VI) problem in that each agent is associated with a local monotone mapping. This problem finds an application in computation of the best equilibrium in nonlinear complementarity problems arising in transportation networks. We develop an iteratively regularized incremental gradient method where at each iteration, agents communicate over a directed cycle graph to update their solution iterates using their local information about the objective and the mapping. The proposed method is single-timescale in the sense that it does not involve any excessive hard-to-project computation per iteration. We derive nonasymptotic agent-wise convergence rates for the suboptimality of the global objective function and infeasibility of the VI constraints measured by a suitably defined dual gap function. Finally, the proposed method appears to be the first fully iterative scheme equipped with iteration complexity that can address distributed optimization problems with VI constraints over cycle graphs.

convergence↗
DOE OSTI2020

Active Learning Accelerated Discovery of Stable Iridium Oxide Polymorphs for the Oxygen Evolution Reaction

The discovery of high-performing and stable materials for sustainable energy applications is a pressing goal in catalysis and materials science. Understanding the relationship between a material’s structure and functionality is an important step in the process, such that viable polymorphs for a given chemical composition need to be identified. Machine-learning-based surrogate models have the potential to accelerate the search for polymorphs that target specific applications. Herein, we report a readily generalizable active-learning (AL) accelerated algorithm for identification of electrochemically stable iridium oxide polymorphs of IrO 2 and IrO 3 . The search is coupled to a subsequent analysis of the electrochemical stability of the discovered structures for the acidic oxygen evolution reaction (OER). Structural candidates are generated by identifying all 956 structurally unique AB2 and AB3 prototypes in existing materials databases (more than 38000). Next, using an active learning approach, we find 196 IrO 2 polymorphs within the thermodynamic amorphous synthesizability limit and reaffirm the global stability of the rutile structure. We find 75 synthesizable IrO 3 polymorphs and report a previously unknown FeF 3 -type structure as the most stable, termed α-IrO 3 . To test the algorithms performance, we compare to a random search of the candidate space and report at least a 2-fold increase in the rate of discovery. Additionally, the AL approach can acquire the most stable polymorphs of IrO 2 and IrO 3 with fewer than 30 density functional theory optimizations. Analysis of the structural properties of the discovered polymorphs reveals that octahedral local coordination environments are preferred for nearly all low-energy structures. Subsequent Pourbaix Ir–H 2 O analysis shows that α-IrO3 is the globally stable solid phase under acidic OER conditions and supersedes the stability of rutile IrO 2 . Calculation of theoretical OER surface activities reveal ideal weaker binding of the OER intermediates on α-IrO 3 than on any other considered iridium oxide. We emphasize that the proposed AL algorithm can be easily generalized to search for any binary metal oxide structure with a defined stoichiometry.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
NASA NTRS2017

Reconfigurable Model Execution in the OpenMDAO Framework

NASA's OpenMDAO framework facilitates constructing complex models and computing their derivatives for multidisciplinary design optimization. Decomposing a model into components that follow a prescribed interface enables OpenMDAO to assemble multidisciplinary derivatives from the component derivatives using what amounts to the adjoint method, direct method, chain rule, global sensitivity equations, or any combination thereof, using the MAUD architecture. OpenMDAO also handles the distribution of processors among the disciplines by hierarchically grouping the components, and it automates the data transfer between components that are on different processors. These features have made OpenMDAO useful for applications in aircraft design, satellite design, wind turbine design, and aircraft engine design, among others. This paper presents new algorithms for OpenMDAO that enable reconfigurable model execution. This concept refers to dynamically changing, during execution, one or more of: the variable sizes, solution algorithm, parallel load balancing, or set of variables-i.e., adding and removing components, perhaps to switch to a higher-fidelity sub-model. Any component can reconfigure at any point, even when running in parallel with other components, and the reconfiguration algorithm presented here performs the synchronized updates to all other components that are affected. A reconfigurable software framework for multidisciplinary design optimization enables new adaptive solvers, adaptive parallelization, and new applications such as gradient-based optimization with overset flow solvers and adaptive mesh refinement. Benchmarking results demonstrate the time savings for reconfiguration compared to setting up the model again from scratch, which can be significant in large-scale problems. Additionally, the new reconfigurability feature is applied to a mission profile optimization problem for commercial aircraft where both the parametrization of the mission profile and the time discretization are adaptively refined, resulting in computational savings of roughly 10% and the elimination of oscillations in the optimized altitude profile.

Hwang, John T.↗
NASA NTRS1989

Optimized orbits for the constellation of Space Station polar platforms

The orbital configuration of the Eos polar-platform constellation is examined to determine if baseline orbital parameters are appropriately specified to meet the mission task of measuring global oceanic mesoscale variability. Using the Nyquist sampling theorem, spatial and temporal measurement requirements are first established for the constellation. Based upon these requirements, the optimum solution indicates that a total of five spacecraft equipped with nadir-pointing altimeters, or two multibeam-equipped platforms, would be required to meet the mesoscale mapping task. Additional analyses were performed on both one- and and two-platform configurations as well as the complete four-platform Eos constellation. The results of these calculations show that the preferred scenario would be to equip all four Eos polar platforms with multibeam altimeters.

Chase, Robert↗
DOE OSTI2020

Collaborative Research: Advancing Arctic Climate Projection Capability at Seasonal to Decadal Scales (Final Technical Report)

The Regional Arctic System Model (RASM) at process resolving configurations has been used to (i) advance understanding of physical processes and feedbacks involved in Arctic amplification and (ii) understand and potentially reduce uncertainty in prediction of arctic climate change at seasonal to decadal scales. RASM consists the atmosphere (Weather and Research Forecasting model, WRF), ocean (Parallel Ocean Program, POP), sea ice (CICE), land hydrology (Variable Infiltration Capacity model, VIC), river routing scheme (RVIC), marine biogeochemistry components and the coupling framework (CPL7). Its domain is pan-Arctic, with the atmosphere and land components configured on a 50-km or 25-km grid and four configurations of the ocean and sea ice components: 1/12°(~9.3km) or 1/48°(~2.4km) and 45 or 60 vertical layers. These RASM configurations have been motivated by the emerging exascale capability for high performance computing to improve model fidelity. The dynamical downscaling of reanalysis allows comparison of RASM results with observations in place and time to: (i) advance system level understanding of physical processes and coupling involved in an event, (ii) optimize model parameter space, (iii) diagnose and reduce model biases and (iv) produce realistic and consistent across all the components initial conditions for predictions and predictability studies, which are all unique capabilities not available in global Earth System Models (ESMs). An evaluation of RASM 1.0 (Cassano et al. 2017) revealed that it had a cold bias over the oceans and a warm bias over land areas due largely to cloud and radiation biases in the model, with too little cloud cover simulated over land and too much cloud cover simulated over sub-polar oceans. This study has motivated an upgrade to WRF version 3.7.1 in RASM and allowed for the inclusion of the radiative impact of convective clouds. A variety of atmospheric physics parameterizations were evaluated against observations (e.g. data from the Arctic Clouds in Summer Experiment (ACSE); Sedlar et al. 2020) to identify an optimal suite of WRF physics options in RASM. The RASM with the optimized WRF physics were used to study the impact of strong mesoscale winds over the ocean around the southern tip of Greenland (DuVivier and Cassano 2016) and their impact on oceanic convection (DuVivier et al. 2017a). Data from the PolarWinds field campaign were used to evaluate WRF boundary layer physics and resolution impacts on the simulation of a Greenland barrier wind event (DuVivier et al. 2017b). The RVIC streamflow routing model has been implemented in RASM to realistically represent high-resolution streamflow processes (Hamman et al. 2017) and to couple the land buoyancy fluxes to the ocean. The RASM-RVIC high-resolution data set of all coastal freshwater fluxes for the Arctic drainage basin and surrounding areas for 1979-2014 was published as a separate product (https://doi.org/10.5281/zenodo.293037). The fidelity of atmospheric momentum transfer to and the response of polar marine Ekman layer in RASM and Community Earth System Model (CESM) was investigated by Roberts et al. (2015). The increased frequency of oceanic flux exchange in CESM, following the RASM guidance, caused a considerable increase in the median inertial ice speed across the Southern Ocean and parts of the Arctic. A comprehensive evaluation of the RASM1.0 atmosphere-ocean-sea ice-land interface was completed by Brunke et al. (2018). RASM was also demonstrated for its capability to simulate extreme events in agreement with observations in space and time (Lee et al. submitted). In particular, the development of three open water events, known as polynyas, have been simulated north of Greenland in February of 2011, 2017 and 2018, in agreement with satellite observations for the past four decades. The optimized RASM sea ice results have been favorably evaluated against satellite observations and a subset of eleven CMIP6 models (Watts et al. submitted). In a complementary project, Jin et al. (2018) have shown that RASM with higher-resolution and new sea-ice processes contributed to lower model errors in sea-ice conditions, concentrations of nutrients and ice algae, in comparison to results from the coarse-resolution (1°) CESM. In two other complementary studies, RASM results were used (i) to explain areas of concentrated use by bowhead whales, the seasonal progression in the use, and the physical environment within those areas (Citta et al. 2015) and (ii) for a synthesis of fall bowhead whales distribution and migration in the Bering-Chukchi-Beaufort (BCB) Sea to investigate whale movements and feeding to the local ocean hydrography and currents (Citta et al. 2018). However, the multi-decadal output from the CESM Large Ensemble yielded unrealistic forcing. Instead, the shorter NCEP CFSv2 9-month forecasts were successfully tested and afforded an increased ensemble size (~30) to demonstrate gains of dynamical downscaling at sub-seasonal to intra-annual time scales. The improved model physics and coupling among RASM model components have yielded more realistic representation of the sea ice cover and consistent across all model components initial conditions. Consequently, RASM demonstrates significant gains compared to simulation of sea ice in the NCEP reanalysis. In addition, RASM 6-month ensemble forecasts yield very realistic sea ice distribution, which demonstrates both significant gains of dynamical downscaling and the continued impact of the initial conditions on forecasts out to 6 months (https://nps.edu/web/rasm/predictions). A follow up study using RASM for dynamical downscaling of the more realistic CESM initialized Decadal Prediction Large Ensemble output is currently ongoing as part of the DOE RGMA HiLAT-RASM project.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2021

Design and Performance of a Smartphone-based Cosmic Ray Observatory

This dissertation examines the feasibility of appropriating the global network of smartphones as an observatory for ultra-high-energy cosmic rays. An application and cloud-based data-acquisition system are first proposed for such an observatory, in which heterogenous devices self-calibrate and server feedback optimizes individual device triggers. Methods of monitoring and manually controlling this network are also examined. Detection efficiencies for cosmic muons, MeV-scale gamma rays, and 120 GeV protons are then measured for this trigger, with which a Monte Carlo pixel model is fine-tuned. Extending this model to cosmic ray showers, the effective area of the observatory for super-GZK primaries is shown to equal that of the Pierre Auger Observatory with a global participation rate under 0.1%, far below initial estimates presented in Ref. [83]. A vastly improved sensitivity to photons and more precise modeling of electrons are primarily responsible for this discrepancy, though more refined models of the combinatorial background may yet lead to more stringent requirements.

Swaney, Jeff↗
DOE OSTI2025

Uncertainty propagation and sensitivity analysis for constrained optimization of nuclear waste vitrification

Abstract The vitrification of high‐level waste (HLW) by heating a mixture of glass‐forming chemicals (GFCs) with the waste can be improved using a constrained optimization problem. This study explores how different uncertainty propagation (UP) methods implemented with the optimization process can affect the glass formulation of nuclear waste glasses. UP is the effort of propagating uncertain inputs through a system to understand and quantify output distributions. Uncertainty intervals are crafted from output distributions to inform the optimization algorithm. UP is often implemented with Monte Carlo (MC) sampling for large nonlinear systems, which can be difficult to implement within a constrained optimization algorithm that requires derivative information. Other UP methods often used for optimization under uncertainty (OUU) can be designed to work within an established constrained optimization framework. Methods of UP are evaluated in this study including iterative sampling approaches, first‐order approximations, and surrogate modeling with machine learning (ML). A method of dimensional reduction based on global sensitivity analysis is introduced to support the UP methods for the large dimensionality of the problem. Analytical UP methods able to achieve similar optimums 10 times faster than the baseline MC approach, and produce 93.9% similar output distributions are reported.

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗
DOE OSTI2020

COHORT: Coordination of Heterogeneous Thermostatically Controlled Loads for Demand Flexibility

Demand flexibility is increasingly important for power grids. Careful coordination of thermostatically controlled loads (TCLs) can modulate energy demand, decrease operating costs, and increase grid resiliency. We propose a novel distributed control framework for the Coordination Of HeterOgeneous Residential Thermostatically controlled loads (COHORT). COHORT is a practical, scalable, and versatile solution that coordinates a population of TCLs to jointly optimize a grid-level objective, while satisfying each TCL’s end-use requirements and operational constraints. To achieve that, we decompose the grid-scale problem into subproblems and coordi- nate their solutions to find the global optimum using the alternating direction method of multipliers (ADMM). The TCLs’ local problems are distributed to and computed in parallel at each TCL, making COHORT highly scalable and privacy-preserving. While each TCL poses combinatorial and non-convex constraints, we characterize these constraints as a convex set through relaxation, thereby making COHORT computationally viable over long planning horizons. After coordination, each TCL is responsible for its own control and tracks the agreed-upon power trajectory with its preferred strategy. In this work, we translate continuous power back to discrete on/off actuation, using pulse width modulation. COHORT is generalizable to a wide range of grid objectives, which we demonstrate through three distinct use cases: generation following, minimizing ramping, and peak load curtailment. In a notable experiment, we validated our approach through a hardware-in-the-loop simulation, including a real-world air conditioner (AC) controlled via a smart thermostat, and simulated instances of ACs modeled after real-world data traces. During the 15-day experimental period, COHORT reduced daily peak loads by an average of 12.5% and maintained comfortable temperatures.

demand response↗
NASA NTRS2000

A Hybrid Satellite-Terrestrial Approach to Aeronautical Communication Networks

Rapid growth in air travel has been projected to continue for the foreseeable future. To maintain a safe and efficient national and global aviation system, significant advances in communications systems supporting aviation are required. Satellites will increasingly play a critical role in the aeronautical communications network. At the same time, current ground-based communications links, primarily very high frequency (VHF), will continue to be employed due to cost advantages and legacy issues. Hence a hybrid satellite-terrestrial network, or group of networks, will emerge. The increased complexity of future aeronautical communications networks dictates that system-level modeling be employed to obtain an optimal system fulfilling a majority of user needs. The NASA Glenn Research Center is investigating the current and potential future state of aeronautical communications, and is developing a simulation and modeling program to research future communications architectures for national and global aeronautical needs. This paper describes the primary requirements, the current infrastructure, and emerging trends of aeronautical communications, including a growing role for satellite communications. The need for a hybrid communications system architecture approach including both satellite and ground-based communications links is explained. Future aeronautical communication network topologies and key issues in simulation and modeling of future aeronautical communications systems are described.

Kerczewski, Robert J.↗
DOE OSTI2026

Biocatalyst discovery and design for plastics deconstruction: A multi‐scale perspective

Plastic waste accumulation poses significant environmental challenges due to a lack of economical solutions for the molecular deconstruction of diverse synthetic polymers. Biological‐based degradation offers promise but is hindered by the crystallinity, hydrophobicity, and additive complexity of plastics, which restrict biocatalyst access and activity. To address these problems, we propose a multi‐scale framework that combines detailed materials characterization, optimization of plastic‐biomolecular interfacial interactions, and enhancement of biocatalytic kinetics to develop effective plastic‐deconstructing enzymes. This approach leverages principles from reaction kinetics, transport and interfacial phenomena, and enzyme engineering to systematically address barriers across diverse plastic types. Our framework aims to accelerate the discovery and optimization of biocatalysts capable of scalable, selective, and efficient deconstruction of plastic waste. These advances hold potential to enable sustainable biological recycling and upcycling pathways, contributing to global efforts in mitigating plastic pollution and promoting circular material economies.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2023

Adaptive Pore Opening to Form Tailored Adsorption Sites in a Cooperatively Flexible Framework Enables Record Inverse Propane/Propylene Separation

A proposed low-energy alternative to the separation of alkanes from alkenes by energy-intensive cryogenic distillation is separation by porous adsorbents. Unfortunately, most adsorbents preferentially take up the desired, high-value major component alkene, requiring frequent regeneration. Adsorbents with inverse selectivity for the minor component alkane would enable the direct production of purified, reagent-grade alkene, greatly reducing global energy consumption. However, such materials are exceedingly rare, especially for propane/propylene separation. Here, we report that through adaptive and spontaneous pore size and shape adaptation to optimize an ensemble of weak noncovalent interactions, the structurally responsive metal-organic framework CdIF-13 (sod-Cd(benzimidazolate) 2 ) exhibits inverse selectivity for propane over propylene with record-setting separation performance under industrially relevant temperature, pressure, and mixture conditions. Powder synchrotron X-ray diffraction measurements combined with first-principles calculations yield atomic-scale insight and reveal the induced fit mechanism of adsorbate-specific pore adaptation and ensemble interactions between ligands and adsorbates. Dynamic column breakthrough measurements confirm that CdIF-13 displays selectivity under mixed-component conditions of varying ratios, with a record measured selectivity factor of α ≈ 3 at 95:5 propylene:propane at 298 K and 1 bar. When sequenced with a low-cost rigid adsorbent, we demonstrated the direct purification of propylene under ambient conditions. In conclusion, this combined atomic-level structural characterization and performance testing firmly establishes how cooperatively flexible materials can be capable of unprecedented separation factors.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
NASA NTRS1986

Equivalent plate analysis of aircraft wing box structures with general planform geometry

A new equilvalent plate analysis formulation is described which is capable of modeling aircraft wing structures with a general planform such as cranked wing boxes. Multiple trapezoidal segments are used to represent such planforms. A Ritz solution technique is used in conjunction with global displacement functions which encompass all the segments. This Ritz solution procedure is implemented efficiently into a computer program so that it can be used by rigorous optimization algorithms for application in early preliminary design. A direct method to interface this structural analysis procedure with aerodynamic programs for use in aeroelastic calculations is described. This equivalent plate analysis procedure is used to calculate the static deflections and stresses and vibration frequencies and modes of an example wing configuration. The numerical results are compared with results from a finite element model of the same configuration to illustrate typical levels of accuracy and computation times resulting from use of this procedure.

Giles, G. L.↗
NASA NTRS1986

Equivalent plate analysis of aircraft wing box structures with general planform geometry

A new equivalent plate analysis formulation is described which is capable of modeling aircraft wing structures with a general planform such as cranked wing boxes. Multiple trapezoidal segments are used to represent such planforms. A Ritz solution technique is used in conjunction with global displacement functions which encompass all the segments. This Ritz solution procedure is implemented efficiently into a computer program so that it can be used by rigorous optimization algorithms for application in early preliminary design. A direct method to interface this structural analysis procedure with aerodynamic programs for use in aeroelastic calculations is described. This equivalent plate analysis procedure is used to calculate the static deflections and stresses and vibration frequencies and modes of an example wing configuration. The numerical results are compared with results from a finite element model of the same configuration to illustrate typical levels of accuracy and computation times resulting from use of this procedure.

Giles, G. L.↗
NASA NTRS1998

Comparison of Response Surface and Kriging Models for Multidisciplinary Design Optimization

In this paper, we compare and contrast the use of second-order response surface models and kriging models for approximating non-random, deterministic computer analyses. After reviewing the response surface method for constructing polynomial approximations, kriging is presented as an alternative approximation method for the design and analysis of computer experiments. Both methods are applied to the multidisciplinary design of an aerospike nozzle which consists of a computational fluid dynamics model and a finite-element model. Error analysis of the response surface and kriging models is performed along with a graphical comparison of the approximations, and four optimization problems m formulated and solved using both sets of approximation models. The second-order response surface models and kriging models-using a constant underlying global model and a Gaussian correlation function-yield comparable results.

Simpson, Timothy W.↗
NASA NTRS2021

Evaluation of Global Surface Water Temperature Data Sets for use in Passive Remote Sensing of Soil Moisture

Inland open water bodies often pose a systematic error source in the passive remote sensing retrievals of soil moisture. Water temperature is a necessary variable used to compute water emissions that is required to be subtracted from satellite observation to yield actual emissions from the land portion, which in turn generates accurate soil moisture retrievals. Therefore, overestimation of soil moisture can often be corrected using concurrent water temperature data in the overall mitigation procedure. In recent years, several data sets of lake water temperature have become available, but their specifications and accuracy have rarely been investigated in the context of passive soil moisture remote sensing on a global scale. For this reason, three lake temperature products were evaluated against in-situ measurements from 2007 to 2011. The data sets include the lake surface water temperature (LSWT) from Global Observatory of Lake Responses to Environmental Change (GloboLakes), the Copernicus Global Land Operations Cryosphere and Water (C-GLOPS), as well as the lake mix-layer temperature (LMLT) from the European Centers for Medium-Range Weather Forecast (ECMWF) ERA5 Land Reanalysis. GloboLakes, C-GLOPS, and ERA5 Land have overall comparable performance with Pearson correlations (R) of 0.87, 0.92 and 0.88 in comparison with in-situ measurements. LSWT products exhibit negative median biases of −0.27 K (GloboLakes) and −0.31 K (C-GLOPS), whereas the median bias of LMLT is 1.56 K. When mapped from their respective native resolutions to a common 9 km Equal-Area Scalable Earth (EASE) Grid 2.0 projection, similar relative performance was observed. LMLT and LSWT data are closer in performance over the 9 km grid cells that exhibit a small range of lake cover fractions (0.05–0.5). Despite comparable relative performance, ERA5 Land shows great advantages in spatial coverage and temporal resolution. In summary, an integrated evaluation on data accuracy, long-term availability, global coverage, temporal resolution, and regular forward processing with modest data latency led us to conclude that LMLT from the ERA5 Land Reanalysis product represents the most optimal path for use in the development of a long-term soil moisture product.

Runze Zhang↗
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↗