Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “level set methods”

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

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

At least 1,549 records · Page 86
All sourcesNASA NTRSDOE OSTI
DOE OSTI2022

Insights into Methodologies and Operational Details of Resource Adequacy Assessment: A Case Study with Application to a Broader Flexibility Framework

Assessing and maintaining resource adequacy (RA) is a core pillar of power systems. However, recent changes in the physical makeup of these systems and the conditions under which these systems must operate have yielded a renewed interest in the methods, metrics, and assumptions that underpin RA assessments. In this paper, we systematically explore a wide range of RA modeling dimensions, including: the objective function and level of operational detail in the underlying model formulation; the quantity (look-ahead) and quality (accuracy) of data that is available for making operational decisions within those models; and the physical configuration of solar photovoltaics (PV) with battery storage hybrid resources. We apply a set of probabilistic RA tools and production cost modeling tools to a realistic test system based loosely on a future Electric Reliability Council of Texas power system dominated by solar PV resources. Under the assumptions of our system and models, we find that multi-stage probabilistic assessments may provide a more robust evaluation of RA by capturing a wider range of operational and system interactions, but this comes at a computational cost of 1-2 orders of magnitude longer run time depending on the specific configuration. In addition, the information on thermal generator availability impacts RA performance by an order of magnitude more than solar resource forecasts, which is driven by the comparatively larger magnitude of thermal outages than solar forecast errors within our test system. Lastly, the flexibility provided by hybrid and other resources can help reduce system load-shedding event frequencies and enable the system to be more robust to inaccurate forecast information, and alternative hybrid inverter sizes can impact RA levels by 1-2 orders of magnitude. Our results point to the importance of a broader flexibility framework to describe the interaction between (1) flexibility "supply" from both physical resource capabilities and operational constraints considered in the modeling, and (2) flexibility "demand" from forecast errors, thermal generator outages, and other sources of uncertainty, as well as their RA impacts. Results are likely sensitive to the system buildout explored; future work could consider additional system configurations and conditions.

ENERGY PLANNING, POLICY, AND ECONOMY,POWER TRANSMI↗
DOE OSTI2025

The latent variable proximal point algorithm for variational problems with inequality constraints

The latent variable proximal point (LVPP) algorithm is a framework for solving infinite-dimensional variational problems with pointwise inequality constraints. The algorithm is a saddle point reformulation of the Bregman proximal point algorithm. At the continuous level, the two formulations are equivalent, but the saddle point formulation is more amenable to discretization because it introduces a structure-preserving transformation between a latent function space and the feasible set. Working in this latent space is much more convenient for enforcing inequality constraints than the feasible set, as discretizations can employ general linear combinations of suitable basis functions, and nonlinear solvers can involve general additive updates. LVPP yields numerical methods with observed mesh-independence for obstacle problems, contact, fracture, plasticity, and others besides; in many cases, for the first time. The framework also extends to more complex constraints, providing means to enforce convexity in the Monge–Ampère equation and handling quasi-variational inequalities, where the underlying constraint depends implicitly on the unknown solution. Here, in this paper, we describe the LVPP algorithm in a general form and apply it to ten problems from across mathematics.

Inequality constraints↗
DOE OSTI2023

Single-Point Extrapolation to the Complete Basis Set Limit through Deep Learning

Machine learning (ML) offers an attractive method for making predictions about molecular systems while circumventing the need to run expensive electronic structure calculations. Once trained on ab initio data, the promise of ML is to deliver accurate predictions of molecular properties that were previously computationally infeasible. In this work, we develop and train a graph neural network model to correct the basis set incompleteness error (BSIE) between a small and large basis set at the RHF and B3LYP levels of theory. Our results show that, when compared to fitting to the total potential, an ML model fitted to correct the BSIE is better at generalizing to systems not seen during training. We test this ability by training on single molecules while evaluating on molecular complexes. We also show that ensemble models yield better behaved potentials in situations where the training data is insufficient. However, even when only fitting to the BSIE, acceptable performance is only achieved when the training data sufficiently resemble the systems one wants to make predictions on. As a result, the test error of the final model trained to predict the difference between the cc-pVDZ and cc-pV5Z potential is 0.184 kcal/mol for the B3LYP density functional, and the ensemble model accurately reproduces the large basis set interaction energy curves on the S66x8 dataset.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2018

Health intelligence: how artificial intelligence transforms population and personalized health

Advances in computational and data sciences for data management, integration, mining, classification, filtering, visualization along with engineering innovations in medical devices have prompted demands for more comprehensive and coherent strategies to address the most fundamental questions in health care and medicine. Theory, methods, and models from artificial intelligence (AI) are changing the health care landscape in clinical and community settings and have already shown promising results in multiple applications in healthcare including, integrated health information systems, patient education, geocoding health data, social media analytics, epidemic and syndromic surveillance, predictive modeling and decision support, mobile health, and medical imaging (e.g. radiology and retinal image analyses). Health intelligence uses tools and methods from artificial intelligence and data science to provide better insights, reduce waste and wait time, and increase speed, service efficiencies, level of accuracy, and productivity in health care and medicine.

Shaban-Nejad, Arash↗
NASA NTRS2010

Near Real-Time Optimal Prediction of Adverse Events in Aviation Data

The prediction of anomalies or adverse events is a challenging task, and there are a variety of methods which can be used to address the problem. In this paper, we demonstrate how to recast the anomaly prediction problem into a form whose solution is accessible as a level-crossing prediction problem. The level-crossing prediction problem has an elegant, optimal, yet untested solution under certain technical constraints, and only when the appropriate modeling assumptions are made. As such, we will thoroughly investigate the resilience of these modeling assumptions, and show how they affect final performance. Finally, the predictive capability of this method will be assessed by quantitative means, using both validation and test data containing anomalies or adverse events from real aviation data sets that have previously been identified as operationally significant by domain experts. It will be shown that the formulation proposed yields a lower false alarm rate on average than competing methods based on similarly advanced concepts, and a higher correct detection rate than a standard method based upon exceedances that is commonly used for prediction.

Martin, Rodney Alexander↗
NASA NTRS2011

Highly Accurate Quartic Force Fields, Vibrational Frequencies, and Spectroscopic Constants for Cyclic and Linear C3H3(+)

High levels of theory have been used to compute quartic force fields (QFFs) for the cyclic and linear forms of the C H + molecular cation, referred to as c-C H + and I-C H +. Specifically the 33 3333 singles and doubles coupled-cluster method that includes a perturbational estimate of connected triple excitations, CCSD(T), has been used in conjunction with extrapolation to the one-particle basis set limit and corrections for scalar relativity and core correlation have been included. The QFFs have been used to compute highly accurate fundamental vibrational frequencies and other spectroscopic constants using both vibrational 2nd-order perturbation theory and variational methods to solve the nuclear Schroedinger equation. Agreement between our best computed fundamental vibrational frequencies and recent infrared photodissociation experiments is reasonable for most bands, but there are a few exceptions. Possible sources for the discrepancies are discussed. We determine the energy difference between the cyclic and linear forms of C H +, 33 obtaining 27.9 kcal/mol at 0 K, which should be the most reliable available. It is expected that the fundamental vibrational frequencies and spectroscopic constants presented here for c-C H + 33 and I-C H + are the most reliable available for the free gas-phase species and it is hoped that 33 these will be useful in the assignment of future high-resolution laboratory experiments or astronomical observations.

Huang, Xinchuan↗
DOE OSTI2019

Characterization of the 2-methylvinoxy radical + O 2 reaction: A focal point analysis and composite multireference study

Vinoxy radicals are involved in numerous atmospheric and combustion mechanisms. High-level theoretical methods have recently shed new light on the reaction of the unsubstituted vinoxy radical with O 2 . The reactions of 1-methylvinoxy radical and 2-methylvinoxy radical with molecular oxygen have experimental high pressure limiting rate constants, k ∞ , 5–7 times higher than that of the vinoxy plus O 2 reaction. In this work, high-level ab initio quantum chemical computations are applied to the 2-methylvinoxy radical plus O 2 system, namely, the formation and isomerization of the 1-oxo-2-propylperoxy radical, the immediate product of O 2 addition to the 2-methylvinoxy radical. Multireference methods were applied to the entrance channel. No barrier to O 2 addition could be located, and more sophisticated treatment of dynamic electron correlation shows that the principal difference between O 2 addition to the vinoxy and 2-methylvinoxy radicals is a larger steric factor for 2-methylvinoxy + O 2 . This is attributed to the favorable interaction between the incoming O 2 molecule and the methyl group of the 2-methylvinoxy radical. Via the focal point approach, energetics for this reaction were determined, in most cases, to chemical accuracy. The coupled-cluster singles, doubles, and perturbative triples [CCSD(T)] correlation energy and Hartree–Fock energies were independently extrapolated to the complete basis set limit. A correction for the effect of higher excitations was computed at the CCSDT(Q)/6-31G level. Corrections for the frozen-core approximation, the Born–Oppenheimer approximation, the nonrelativistic approximation, and the zero-point vibrational energy were included. From the 1-oxo-2-propylperoxy radical, dissociation to reactants is competitive with the lowest energy isomerization pathway. The lowest energy isomerization pathway ultimately forms acetaldehyde, CO, and ·OH as the final products.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
NASA NTRSDate not supplied

FUTURE LEVELS OF AEROSOLS & RADIATIVE FORCING OVER THE ARCTIC

The Arctic Monitoring and Assessment Programme (AMAP) is currently assessing the impacts of Short-Lived Climate Forcers (SLCF) on Arctic climate and air quality. In support of the assessment, we used the NASA Goddard Institute of Space Sciences (GISS) Earth System Model (modelE2.1), with prescribed sea surface temperature and sea-ice fraction, to simulate SLCF concentrations globally between 1995 and 2015. Two simulations were conducted, using the One-Moment Aerosol (OMA) and the Multiconfiguration Aerosol TRacker of mIXing state (MATRIX) aerosol modules. OMA is a mass-based scheme in which aerosols are assumed to remain externally mixed and have a prescribed and constant size distribution, while MATRIX is an aerosol microphysics scheme based on the quadrature method of moments, which is able to explicitly simulate the mixing state of aerosols. Anthropogenic emissions from the ECLIPSE v6b emissions database were used, along with emissions from aircrafts and open biomass burning from the Coupled Model Intercomparison Project Phase 6 (CMIP6), while the natural emissions of sea salt, DMS, isoprene and dust are calculated interactively. The simulated monthly surface concentrations of sulfate (SO4), black carbon (BC), organic carbon (OA), and ozone (O3) are compared with observations from a set of Arctic stations, extracted from the EBAS and IMPROVE databases, as well as a few additional locations. Simulated aerosol optical depths (AOD) are also compared with Advanced Very-High Resolution Radiometer (AVHRR). The study will present the evaluation of the modelE2.1 in simulating SLCF levels over the Arctic using different aerosol schemes, along with observed and simulated trends of SLCFs over the Arctic between 1995 and 2015.

Aersosols↗
DOE OSTI2015

Plackett-Burman experimental design to facilitate syntactic foam development

The use of an eight-experiment Plackett–Burman method can assess six experimental variables and eight responses in a polysiloxane-glass microsphere syntactic foam. The approach aims to decrease the time required to develop a tunable polymer composite by identifying a reduced set of variables and responses suitable for future predictive modeling. The statistical design assesses the main effects of mixing process parameters, polymer matrix composition, microsphere density and volume loading, and the blending of two grades of microspheres, using a dummy factor in statistical calculations. Responses cover rheological, physical, thermal, and mechanical properties. The cure accelerator content of the polymer matrix and the volume loading of the microspheres have the largest effects on foam properties. These factors are the most suitable for controlling the gel point of the curing foam, and the density of the cured foam. The mixing parameters introduce widespread variability and therefore should be fixed at effective levels during follow-up testing. Some responses may require greater contrast in microsphere-related factors. As a result, compared to other possible statistical approaches, the run economy of the Plackett–Burman method makes it a valuable tool for rapidly characterizing new foams.

36 MATERIALS SCIENCE↗
DOE OSTI2014

Bayesian Proteoform Modeling Improves Protein Quantification of Global Proteomic Measurements

As the capability of mass spectrometry-based proteomics has matured, tens of thousands of peptides can be measured simultaneously, which has the benefit of offering a systems view of protein expression. However, a major challenge is that with an increase in throughput, protein quantification estimation from the native measured peptides has become a computational task. A limitation to existing computationally-driven protein quantification methods is that most ignore protein variation, such as alternate splicing of the RNA transcript and post-translational modifications or other possible proteoforms, which will affect a significant fraction of the proteome. The consequence of this assumption is that statistical inference at the protein level, and consequently downstream analyses, such as network and pathway modeling, have only limited power for biomarker discovery. Here, we describe a Bayesian model (BP-Quant) that uses statistically derived peptides signatures to identify peptides that are outside the dominant pattern, or the existence of multiple over-expressed patterns to improve relative protein abundance estimates. It is a research-driven approach that utilizes the objectives of the experiment, defined in the context of a standard statistical hypothesis, to identify a set of peptides exhibiting similar statistical behavior relating to a protein. This approach infers that changes in relative protein abundance can be used as a surrogate for changes in function, without necessarily taking into account the effect of differential post-translational modifications, processing, or splicing in altering protein function. We verify the approach using a dilution study from mouse plasma samples and demonstrate that BP-Quant achieves similar accuracy as the current state-of-the-art methods at proteoform identification with significantly better specificity. BP-Quant is available as a MatLab ® and R packages at https://github.com/PNNL-Comp-Mass-Spec/BP-Quant.

Protein Quantitation, proteomics↗
DOE OSTI2021

Computational design and analysis of modular cells for large libraries of exchangeable product synthesis modules

Microbial metabolism can be harnessed to produce a large library of useful chemicals from renewable resources such as plant biomass. However, it is laborious and expensive to create microbial biocatalysts to produce each new product. To tackle this challenge, we have recently developed modular cell (ModCell) design principles that enable rapid generation of production strains by assembling a modular (chassis) cell with exchangeable production modules to achieve overproduction of target molecules. Previous computational ModCell design methods are limited to analyze small libraries of around 20 products. In this study, we developed a new computational method, named ModCell-HPC, that can design modular cells for large libraries with hundreds of products with a highly-parallel and multi-objective evolutionary algorithm and enable us to elucidate modular design properties. We demonstrated ModCell-HPC to design Escherichia coli modular cells towards a library of 161 endogenous production modules. From these simulations, we identified E. coli modular cells with few genetic manipulations that can produce dozens of molecules in a growth-coupled manner with different types of fermentable sugars. These designs revealed key genetic manipulations at the chassis and module levels to accomplish versatile modular cells, involving not only in the removal of major by-products but also modification of branch points in the central metabolism. We further found that the effect of various sugar degradation on redox metabolism results in lower compatibility between a modular cell and production modules for growth on pentoses than hexoses. To better characterize the degree of compatibility, we developed a method to calculate the minimal set cover, identifying that only three modular cells are all needed to couple with up 85 compatible production modules. By determining the unknown compatibility contribution metric, we further elucidated the design features that allow an existing modular cell to be re-purposed towards production of new molecules. Altogether, ModCell-HPC is a useful tool for understanding modularity of biological systems and guiding more efficient and generalizable design of modular cells that help reduce research and development cost in biocatalysis.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2016

HNS{sup +} and HSN{sup +} cations: Electronic states, spin-rovibronic spectroscopy with planetary and biological implications

Ab initio methods in conjunction with a large basis set are used to compute the potential energy surfaces of the 12 lowest electronic states of the HNS{sup +} and HSN{sup +} isomeric forms. These potentials are used in discussions of the metastability of these cations and plausible mechanisms for the H{sup +}/H + SN{sup +}/SN, S/S{sup +} + NH{sup +}/NH, N/N{sup +} + SH{sup +}/SH ion-molecule reactions. Interestingly, the low rovibrational levels of HSN{sup +}(1{sup 2}A″) and HNS{sup +}(1{sup 2}A″) electronically excited ions are predicted to be long-lived. Both ions are suggested to be a suitable candidate for light-sensitive NO{sup ⋅} donor in vivo and as a possible marker for the detection of intermediates in nitrites + H{sub 2}S reactions at the cellular level. The full spin rovibronic levels of HNS{sup +} are presented, which may assist in the experimental identification of HNS{sup +} and HSN{sup +} ions and in elucidating their roles in astrophysical and biological media.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2024

Singular Perturbation-Based Large-Signal Order Reduction of Microgrids for Stability and Accuracy Synthesis With Control

The increasing penetration of distributed energy resources (DERs) highlights the growing importance of microgrids (MGs) in enhancing power system reliability. Employing electromagnetic transient (EMT) analysis in MGs becomes crucial for controlling the rapid transients. However, this requires an accurate but high-order model of power electronics and their underlying control loops, complexifying the stability analysis from the viewpoint of a higher control level. To overcome these challenges, this paper proposes a large-signal order reduction (LSOR) method for MGs with considerations of external control inputs and the detailed dynamics of underlying control levels based on singular perturbation theory (SPT). Specially, we innovatively proposed and strictly proved a general stability and accuracy assessment theorem that allows us to analyze the dynamic stability of a full-order nonlinear system by only leveraging our derived reduced-order model (ROM) and boundary layer model (BLM). Furthermore, this theorem furnishes a set of conditions that determine the accuracy of the developed ROM. Lastly, by embedding such a theorem into the SPT, we propose a novel LSOR approach with guaranteed accuracy and stability analysis equivalence. Case studies are conducted on MG systems to show the effectiveness of the proposed approach.

24 POWER TRANSMISSION AND DISTRIBUTION↗
DOE OSTI2013

Comparative dose evaluations between XVI and OBI cone beam CT systems using Gafchromic XRQA2 film and nanoDot optical stimulated luminescence dosimeters

Purpose: To investigate the effect of energy (kVp) and filters (no filter, half Bowtie, and full Bowtie) on the dose response curves of the Gafchromic XRQA2 film and nanoDot optical stimulated luminescence dosimeters (OSLDs) in CBCT dose fields. To measure surface and internal doses received during x-ray volume imager (XVI) (Version R4.5) and on board imager (OBI) (Version 1.5) CBCT imaging protocols using these two types of dosimeters. Methods: Gafchromic XRQA2 film and nanoDot OSLD dose response curves were generated at different kV imaging settings used by XVI (software version R4.5) and OBI (software version 1.5) CBCT systems. The settings for the XVI system were: 100 kVp/F0 (no filter), 120 kVp/F0, and 120 kVp/F1 (Bowtie filter), and for the OBI system were: 100 kVp/full fan, 125 kVp/full fan, and 125 kVp/half fan. XRQA2 film was calibrated in air to air kerma levels between 0 and 11 cGy and scanned using reflection scanning mode with the Epson Expression 10000 XL flat-bed document scanner. NanoDot OSLDs were calibrated on phantom to surface dose levels between 0 and 14 cGy and read using the inLight{sup TM} MicroStar reader. Both dosimeters were used to measure in field surface and internal doses in a male Alderson Rando Phantom. Results: Dose response curves of XRQA2 film and nanoDot OSLDs at different XVI and OBI CBCT settings were reported. For XVI system, the surface dose ranged between 0.02 cGy in head region during fast head and neck scan and 4.99 cGy in the chest region during symmetry scan. On the other hand, the internal dose ranged between 0.02 cGy in the head region during fast head and neck scan and 3.17 cGy in the chest region during chest M20 scan. The average (internal and external) dose ranged between 0.05 cGy in the head region during fast head and neck scan and 2.41 cGy in the chest region during chest M20 scan. For OBI system, the surface dose ranged between 0.19 cGy in head region during head scan and 4.55 cGy in the pelvis region during spot light scan. However, the internal dose ranged between 0.47 cGy in the head region during head scan and 5.55 cGy in the pelvis region during spot light scan. The average (internal and external) dose ranged between 0.45 cGy in the head region during head scan and 3.59 cGy in the pelvis region during spot light scan. Both Gafchromic XRQA2 film and nanoDot OSLDs gave close estimation of dose (within uncertainties) in many cases. Though, discrepancies of up to 20%-30% were observed in some cases. Conclusions: Dose response curves of Gafchromic XRQA2 film and nanoDot OSLDs indicated that the dose responses of these two dosimeters were different even at the same photon energy when different filters were used. Uncertainty levels of both dosimetry systems were below 6% at doses above 1 cGy. Both dosimetry systems gave almost similar estimation of doses (within uncertainties) in many cases, with exceptions of some cases when the discrepancy was around 20%-30%. New versions of the CBCT systems (investigated in this study) resulted in lower imaging doses compared with doses reported on earlier versions in previous studies.

61 RADIATION PROTECTION AND DOSIMETRY↗
DOE OSTI2025

Search for a light charged Higgs boson in $t \rightarrow H^{\pm } b$ decays, with $H^{\pm } \rightarrow cs$, in $pp$ collisions at $\sqrt{s}={13}\hbox { TeV}$ with the ATLAS detector

A search for a light charged Higgs boson produced in decays of the top quark, $t \rightarrow H^{\pm } b$ with $H^{\pm } \rightarrow cs$, is presented. This search targets the production of top-quark pairs $t\bar{t} \rightarrow WbH^{\pm } b$, with $W \rightarrow ℓv(ℓ = e, μ)$, resulting in a lepton-plus-jets final state characterised by an isolated electron or muon and at least four jets. The search exploits b-quark and c-quark identification techniques as well as multivariate methods to suppress the dominant $t\bar{t}$ background. The data analysed correspond to 140 fb -1 of $pp$ collisions at $\sqrt{s}$ = 13 TeV recorded with the ATLAS detector at the LHC between 2015 and 2018. Observed (expected) 95% confidence-level upper limits on the branching fraction $\mathscr{B}(t \rightarrow H^{\pm } b)$, assuming $\mathscr{B}(t \rightarrow Wb) + \mathscr{B}(t \rightarrow H^{\pm }(\rightarrow cs)b$, are set between 0.066% (0.077%) and 3.6% (2.3%) for a charged Higgs boson with a mass between 60 and 168 GeV.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2012

Measurement of the Single Top Quark Cross Section in the Lepton Plus Jets Final State in Proton-Antiproton Collisions at a Center of Mass Energy of 1.96 TeV Using the CDF II Detector

We present a measurement of the single top quark cross section in the lepton plus jets final state using an integrated luminosity corresponding to 7.5 fb -1 of p\bar p collision data collected by the Collider Detector at Fermilab. The single top candidate events are identified by the signature of a charged lepton, large missing transverse energy, and two or three jets with at least one of them identified as originating from a bottom quark. A new Monte Carlo generator POWHEG is used to model the single top quark production processes, which include s-channel, t-channel, and Wt-channel. A neural network multivariate method is exploited to discriminate the single top quark signal from the comparatively large backgrounds. We measure a single top production cross section of $3.04^{+0.57}_{-0.53} (\mathrm{stat.~+~syst.})$ pb assuming $m_{\rm top}=172.5$~GeV/$c^2$. In addition, we extract the CKM matrix element value $|V_{tb}|=0.96\pm 0.09~(\mathrm{stat.~+~syst.})\ ± 0.05~(\mathrm{theory})$ and set a lower limit of $|V_{tb}|>0.78$ at the 95% credibility level.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
NASA NTRS1992

The numerical dynamic for highly nonlinear partial differential equations

Problems associated with the numerical computation of highly nonlinear equations in computational fluid dynamics are set forth and analyzed in terms of the potential ranges of spurious behaviors. A reaction-convection equation with a nonlinear source term is employed to evaluate the effects related to spatial and temporal discretizations. The discretization of the source term is described according to several methods, and the various techniques are shown to have a significant effect on the stability of the spurious solutions. Traditional linearized stability analyses cannot provide the level of confidence required for accurate fluid dynamics computations, and the incorporation of nonlinear analysis is proposed. Nonlinear analysis based on nonlinear dynamical systems complements the conventional linear approach and is valuable in the analysis of hypersonic aerodynamics and combustion phenomena.

Lafon, A.↗
NASA NTRS2005

Distributed Evaluation Functions for Fault Tolerant Multi-Rover Systems

The ability to evolve fault tolerant control strategies for large collections of agents is critical to the successful application of evolutionary strategies to domains where failures are common. Furthermore, while evolutionary algorithms have been highly successful in discovering single-agent control strategies, extending such algorithms to multiagent domains has proven to be difficult. In this paper we present a method for shaping evaluation functions for agents that provide control strategies that both are tolerant to different types of failures and lead to coordinated behavior in a multi-agent setting. This method neither relies of a centralized strategy (susceptible to single point of failures) nor a distributed strategy where each agent uses a system wide evaluation function (severe credit assignment problem). In a multi-rover problem, we show that agents using our agent-specific evaluation perform up to 500% better than agents using the system evaluation. In addition we show that agents are still able to maintain a high level of performance when up to 60% of the agents fail due to actuator, communication or controller faults.

Agogino, Adrian↗