SEARCH · Engineering Papers

Results for “reliability metrics”

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.

SourceResearch connectionFrom yearThrough year

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

At least 577 records · Page 32

All sources NASA NTRS DOE OSTI

NASA NTRS1984

The Sternberg Task as a Workload Metric in Flight Handling Qualities Research

The objective of this research was to determine whether the Sternberg item-recognition task, employed as a secondary task measure of spare mental capacity for flight handling qualities (FHQ) simulation research, could help to differentiate between different flight-control conditions. FHQ evaluations were conducted on the Vertical Motion Simulator at Ames Research Center to investigate different primary flight-control configurations, and selected stability and control augmentation levels for helicopers engaged in low-level flight regimes. The Sternberg task was superimposed upon the primary flight-control task in a balanced experimental design. The results of parametric statistical analysis of Sternberg secondary task data failed to support the continued use of this task as a measure of pilot workload. In addition to the secondary task, subjects provided Cooper-Harper pilot ratings (CHPR) and responded to a workload questionnaire. The CHPR data also failed to provide reliable statistical discrimination between FHQ treatment conditions; some insight into the behavior of the secondary task was gained from the workload questionnaire data.

Hemingway, J. C.↗

DOE OSTI2024

Distribution Grid Impact Study in Highland Park, Michigan: Understanding Rooftop Solar, Behind-the-Meter Energy Storage, Electric Vehicle Charging, and Building Electrification [Slides]

Through the Communities LEAP (Local Energy Action Program) Pilot, the National Renewable Energy Laboratory (NREL) engaged the Highland Park Stakeholder Coalition to scope four technical assistance work areas to address their energy needs and goals. This slide deck addresses the highlighted tasks under work area "B" related to policy analysis, due diligence, and case studies for removing barriers and providing best practices for local clean energy development. Highland Park community members face frequent, long-duration power interruptions due largely to the aging distribution system serving the area and the legacy design standards used in its construction. While degrading physical infrastructure such as poles, crossarms, and transformers can result in this substandard reliability, another notable characteristic of this legacy system is the lower, 4.8 kV, voltage class. This is a legacy construction standard which many utilities, DTE included, are phasing out in favor of higher, 15 or 25 kV, voltage classes instead. The existing 4.8 kV distribution system serving Highland Park may limit significant adoptions of clean-energy technologies like high percentages of building electrification or electric vehicle adoption. The following analysis seeks to quantify these limitations under a variety of clean-energy technology adoption scenarios. It compares the overall system risks of the present system to those of a hypothetical, upgraded 13.2 kV system, using NREL-developed risk metrics and offers upgrade cost considerations. The legacy 4.8 kV voltage class serving Highland park is not, as we have modeled it, a major limitation to the widespread adoption of cost-optimal rooftop solar and/or behind-the-meter energy storage. Within our modeling framework, these technologies namely impact secondary, low-voltage assets, which may be remedied without the need for a system-wide upgrade to a 13.2 kV voltage class. Our model of the current 4.8kV system indicates it is not capable of supporting community-wide electrification efforts. Widespread building electrification, and the resulting large increase in wintertime load, dramatically increases the prevalence of voltage violations and thermal overloading on the current 4.8 kV distribution system. These impacts illustrate the need for system-wide upgrades to a 13.2 kV voltage class to accommodate these technologies. Low to Moderate DER adoption does not adversely impact the grid but does improve undervoltage and asset overloading issues. However, these benefits are insufficient to defer grid upgrades. Higher DER penetration is shown to increase overvoltage and asset overloading in future electrification scenarios.

29 ENERGY PLANNING, POLICY, AND ECONOMY↗

NASA NTRS2024

Quantification of Uncertainty and Risk Sensitivity for Safety of Emerging Operations

The growing need to develop and deploy small unmanned aerial vehicles (sUAVs) for various applications in the airspace necessitates reliable tools to accurately predict the flight trajectories of the sUAVs. The knowledge of the predicted trajectories help decision makers anticipate potential conflict, assess the risk, and take appropriate risk mitigation actions. In addition, uncertainties in vehicle models, weather, and controller action further highlights the need for reliable prediction tools. In this project, the application of mixed sparse grid-based quadrature and generalized polynomial chaos(gPC) expansion method for uncertainty quantification and collision assessment in air traffic consisting of fixed-wing small unmanned aerial vehicles (sUAV) was studied. From the results obtained, it can be concluded that this provides a reliable framework to carry out quantitative conflict assessment in an unmanned air traffic, which when employed, can improve the functionalities of the unmanned traffic management system. It was observed that the results from the gPC expansion framework developed in the project can be utilized to conduct rapid probabilistic collision assessment for near real-time unmanned traffic management in the airspace. From the vehicle models, position updates, and wind-field data, a priori gPC based 3-σcon-fidence ellipses can provide estimates of potential conflict at some future instants. The computational costs scaled linearly when the uncertain inputs were fewer. Further, the largest allowable distribution of para-metric uncertainties that leads to the smallest risk of collision in traffic of small unmanned aerial vehicles could be calculated. The time of closest approach between two sUAVs can be established paving way for development of proactive mitigation strategies. The separation between the sUAVs was found to be most significantly affected by uncertainties in the maximum available thrusts, zero-lift drag coefficients, and wing planform areas of the sUAVs. The study of uncertain wind-fields indicated that a heterogeneous traffic mix resulted in an increased probability of conflict. Increased measurement update rate reduced the uncertain-ties in the trajectories of the vehicles, further reducing the probability of conflict but rapid updates of all vehicles in the airspace poses a stringent communication limitation. The gPC framework also provided the means to analyze vehicle impact (crash region) due to loss of control resulting from actuator failure in sUAS traffic, essentially to predict impact and crash zones for representative vehicles. The predicted regions when compared with non-participant density, provides a means to develop an early mitigation strategy, should the sUAV detect an imminent actuator failure.

Rajnish Bhusal↗

DOE OSTI2021

Managing Solar Photovoltaic Integration in the Western United States: Resource Adequacy Considerations

This study examines the impact of reserve margin-based reliability assessment, as commonly used in capacity expansion models, on planning resource-adequate power systems under high penetrations of solar photovoltaics (PV). As a generation resource, PV is operationally different from the conventional dispatchable resources for which most capacity expansion models were designed. The question this study attempts to answer is whether large amounts of PV on a system (in this case, the Western Interconnection of North America) would bias the results of conventional reserve margin-based capacity expansion modeling towards an over- or under-provisioning of resource adequacy. This analysis used NREL’s Resource Planning Model (RPM) for capacity expansion modeling and NREL’s Probabilistic Resource Adequacy Suite (PRAS) for resource adequacy assessment. RPM uses a reserve margin requirement to enforce resource adequacy. PRAS, a collection of tools for studying the resource adequacy of power systems and the adequacy contributions of individual resources on a probabilistic basis, was used to compute multiple resource adequacy metrics across a number of simulated scenarios and system representations with differing regional detail. In all cases, including high PV penetrations (up to 33% annual generation from PV, interconnection-wide), RPM was able to produce resource-adequate systems as measured by normalized expected unserved energy and loss-of-load expectation results from PRAS. The accuracy of reserve margin approaches depends heavily on the underlying assumptions informing the capacity credit assigned to variable and energy-limited resources, particularly when such resources are abundant in the modeled system. RPM’s standard methodology for estimating variable and flexible resources’ capacity contributions, which is based on the top 100 hours of net load, did not appear to systematically undervalue or overvalue variable generation relative to a more rigorous equivalent firm capacity assessment using PRAS, although both over- and under-valuations were observed in specific scenarios. In the worst cases, the top 100 hour method underestimated the equivalent firm capacity of PV by two percentage points, and overestimated the equivalent firm capacity of PV by five percentage points. Calculating capacity contributions based on the top 10 hours of net load systematically underestimated equivalent firm capacities at more modest PV penetrations, but was often a better approximation of equivalent firm capacity than the existing 100-hour approach in scenarios with higher PV penetrations.

14 SOLAR ENERGY↗

DOE OSTI2021

Informing nuclear physics via machine learning methods with differential and integral experiments

Information from differential nuclear-physics experiments and theory is often too uncertain to accurately define nuclear-physics observables such as cross sections or energy spectra. Integral experimental data, representing the applications of these observables, are often more precise but depend simultaneously on too many of them to unambiguously identify issues in the observable with human expert analysis alone. Here, we explore how we can leverage physics knowledge gained from differential experimental data, nuclear theory, integral experiments, and neutron-transport calculations to better understand nuclear-physics observables in the context of the application area represented by integral experiments. We support this task with machine-learning methods to discern trends in a large amount of convoluted data. Differential and integral information was used in an analysis augmented by the random forest and the Shapley additive explanations metric. We chose as an application area one that is represented by criticality measurements and pulsed-sphere neutron-leakage spectra. We show one representative example ( 241 Pu fission observables) where the combination of differential and integral information allowed to resolve issues in data representing these observables. As a starting point, the machine learning (ML) algorithms highlighted several observables as leading potentially to bias in simulating integral experiments. Differential information, paired with sensitivity to integral quantities, allowed us then to pinpoint one specific observable ( 241 Pu fission cross section) as the main driver of bias. The comparison to integral experiments, on the other hand, allowed us to indicate a likely reliable experiment among several discrepant ones for this observables. In other cases (e.g., 239 Pu observables), we were not able to resolve the confounding introduced by integral experiments but instead highlighted the need for targeted new experiments and theory developments to better constrain the nuclear-physics space for the application area represented by integral experiments. We were able to combine information from differential experimental data, nuclear-physics theory, integral experiments, and neutron-transport simulations of the latter experiments with the help of the random forest algorithm and expert judgment. This combination of knowledge allows to improve our description of nuclear-physics observables as applied to a particular application area.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

NASA NTRS1998

Reusable Rocket Engine Operability Modeling and Analysis

This paper describes the methodology, model, input data, and analysis results of a reusable launch vehicle engine operability study conducted with the goal of supporting design from an operations perspective. Paralleling performance analyses in schedule and method, this requires the use of metrics in a validated operations model useful for design, sensitivity, and trade studies. Operations analysis in this view is one of several design functions. An operations concept was developed given an engine concept and the predicted operations and maintenance processes incorporated into simulation models. Historical operations data at a level of detail suitable to model objectives were collected, analyzed, and formatted for use with the models, the simulations were run, and results collected and presented. The input data used included scheduled and unscheduled timeline and resource information collected into a Space Transportation System (STS) Space Shuttle Main Engine (SSME) historical launch operations database. Results reflect upon the importance not only of reliable hardware but upon operations and corrective maintenance process improvements.

Christenson, R. L.↗

DOE OSTI2023

Deep neural network uncertainty quantification for LArTPC reconstruction

We evaluate uncertainty quantification (UQ) methods for deep learning applied to liquid argon time projection chamber (LArTPC) physics analysis tasks. As deep learning applications enter widespread usage among physics data analysis, neural networks with reliable estimates of prediction uncertainty and robust performance against overconfidence and out-of-distribution (OOD) samples are critical for their full deployment in analyzing experimental data. While numerous UQ methods have been tested on simple datasets, performance evaluations for more complex tasks and datasets are scarce. Here we assess the application of selected deep learning UQ methods on the task of particle classification using the PiLArNet monte carlo 3D LArTPC point cloud dataset. We observe that UQ methods not only allow for better rejection of prediction mistakes and OOD detection, but also generally achieve higher overall accuracy across different task settings. We assess the precision of uncertainty quantification using different evaluation metrics, such as distributional separation of prediction entropy across correctly and incorrectly identified samples, receiver operating characteristic curves (ROCs), and expected calibration error from observed empirical accuracy. We conclude that ensembling methods can obtain well calibrated classification probabilities and generally perform better than other existing methods in deep learning UQ literature.

47 OTHER INSTRUMENTATION↗

DOE OSTI2021

In-Home Mobility Frequency and Stability in Older Adults Living Alone With or Without MCI: Introduction of New Metrics

Background: Older adults spend a considerable amount of time inside their residences; however, most research investigates out-of-home mobility and its health correlates. We measured indoor mobility using room-to-room transitions, tested their psychometric properties, and correlated indoor mobility with cognitive and functional status. Materials and Methods: Community-dwelling older adults living alone ( n = 139; age = 78.1 ± 8.6 years) from the Oregon Center for Aging & Technology (ORCATECH) and Minority Aging Research Study (MARS) were included in the study. Two indoor mobility features were developed using non-parametric parameters (frequency; stability): Indoor mobility frequency (room-to-room transitions/day) was detected using passive infrared (PIR) motion sensors fixed on the walls in four geographic locations (bathroom; bedroom; kitchen; living room) and using door contact sensors attached to the egress door in the entrance. Indoor mobility stability was estimated by variances of number of room-to-room transitions over a week. Test-retest reliability (Intra-class coefficient, ICC) and the minimal clinically important difference (MCID) defined as the standard error of measurement (SEM) were generated. Generalized estimating equations models related mobility features with mild cognitive impairment (MCI) and functional status (gait speed). Results: An average of 206 days (±127) of sensor data were analyzed per individual. Indoor mobility frequency and stability showed good to excellent test-retest reliability (ICCs = 0.91[0.88–0.94]; 0.59[0.48–0.70]). The MCIDs of mobility frequency and mobility stability were 18 and 0.09, respectively. On average, a higher indoor mobility frequency was associated with faster gait speed (β = 0.53, p = 0.04), suggesting an increase of 5.3 room-to-room transitions per day was associated with an increase of 10 cm/s gait speed. A decrease in mobility stability was associated with MCI (β = −0.04, p = 0.03). Discussion: Mobility frequency and stability in the home are clinically meaningful and reliable features. Pervasive-sensing systems deployed in homes can objectively reveal cognitive and functional status in older adults who live alone.

Wu, Chao-Yi↗

NASA NTRSDate not supplied

A Historical Review of Logistics Mass and Crew Time Demands for ISS Operations

Following over 20 years of continuously crewed operations on the International Space Station (ISS), NASA is planning to return to the Moon and eventually send humans to Mars. ISS operations provide vital data to inform mission analysts as NASA prepares for longer and more complex missions with increased mission endurance. Endurance, defined as crewed operating time between cargo deliveries (or crew launch and return to Earth), is an important metric when analyzing mission needs. NASA is developing architectures to support sustained deep-space habitats in cislunar space, the lunar surface, Mars transit, and the surface of Mars. Unlike the ISS, these systems will not be continuously crewed, and unlike the Space Shuttle, these systems will not return to Earth for regular refurbishment between missions. Lunar systems will routinely go through long uncrewed periods between crewed missions. The systems on board will need to survive these dormancy periods with no crew present to provide maintenance. Mars systems will experience significantly longer endurance than past experience. Additionally, the inability to have quick aborts to return to Earth increases the need for system reliability, redundancy, and maintainability, as well as plans for contingency operations. This paper examines the historical logistics and crew time demand for ISS operations and mission objectives and provides an overview of missions to the ISS over its operating history, the mass and items delivered with the missions, and the crew time spent during missions. These parameters provide insight and valuable data to inform logistics and crew time estimates for future long-endurance crewed exploration missions.

Crew Logistics↗

NASA NTRSDate not supplied

A Historical Review of Logistics Mass and Crew Time Demands for ISS Operations

Crew Logistics↗

DOE OSTI2025

Challenges in predicting protein-protein interactions of understudied viruses: Arenavirus-human interactions

Understanding protein-protein interactions (PPIs) between viruses and host organisms is crucial for uncovering infection mechanisms and identifying potential therapeutic targets. The ability to generalize PPI predictive models across understudied viruses presents a significant challenge. In this work, we use arenavirus-human PPIs to illustrate the difficulties associated with model generalization, which are compounded by a lack of both positive and negative data. We employ a Transfer Learning approach to investigate arenavirus-human PPIs by utilizing models trained on better-studied virus-human and human-human PPIs. Additionally, we curate and assess four types of negative sampling datasets to evaluate their impact on model performance. Despite the overall high accuracies (93–99 %) and AUPRC scores (0.8–0.9) appearing promising, further analysis indicates that these performance metrics can be misleading due to data leakage, data bias, and overfitting, especially concerning under-represented viral proteins. We reveal these gaps and assess the impact of data imbalance using standard k-fold cross-validation and Independent Blind Testing with a Balanced Dataset, resulting in a drop in accuracy below 50 %. We propose a viral protein-specific evaluation framework that categorizes viral proteins into majority and minority classes based on their representation in the dataset, enabling comparison of model performance across these groups using balanced accuracies. This framework offers a more robust evaluation of model generalizability, addressing biases inherent in standard evaluation techniques and paving the way for more reliable PPI prediction models for understudied viruses.

59 BASIC BIOLOGICAL SCIENCES↗

DOE OSTI2026

Are Capacity and Energy Loss Equivalent Metrics for Battery Aging Reporting?

Battery aging in research publications and manufacturer specification sheets for individual cells is commonly reported as capacity (Ah) versus cycle number. However, the key measured quantity in battery-powered devices is energy (Wh), which is derived from integrating capacity with voltage. In this work, we compare the rate of capacity and energy loss across a wide range of Li-ion single-cell cycling studies with different positive electrode chemistries, charge–discharge rates, and temperatures. We find that the relative rate of discharge energy loss varies with cycling conditions. For many cells cycled under moderate conditions, the rate of discharge energy fade is only slightly faster than the rate of discharge capacity fade. However, some cells demonstrated up to a 15% decline in cycle count when 80% energy retention rather than 80% capacity retention was used as the end-of-life metric. These results highlight the importance of reporting cell aging based on energy fade to avoid overestimating battery lifetime in full systems.

batteries↗

NASA NTRS1994

Nuclear electric propulsion: A better, safer, cheaper transportation system for human exploration of Mars

NASA has completed a preliminary mission and systems study of nuclear electric propulsion (NEP) systems for 'split-sprint' human exploration and related robotic cargo missions to Mars. This paper describes the study, the mission architecture selected, the NEP system and technology development needs, proposed development schedules, and estimated development costs. Since current administration policy makers have delayed funding for key technology development activities that could make Mars exploration missions a reality in the near future, NASA will have time to evaluate various alternate mission options, and it appears prudent to ensure that Mars mission plans focus on astronaut and mission safety, while reducing costs to acceptable levels. The split-sprint nuclear electric propulsion system offers trip times comparable to nuclear thermal propulsion (NTP) systems, while providing mission abort opportunities that are not possible with 'reference' mission architectures. Thus, NEP systems offer short transit times for the astronauts, reducing the exposure of the crew to intergalactic cosmic radiation. The high specific impulse of the NEP system, which leads to very low propellant requirements, results in significantly lower 'initial mass in low earth orbit' (IMLEO). Launch vehicle packaging studies show that the NEP system can be launched, assembled, and deployed, with about one less 240-metric-ton heavy lift launch vehicle (HLLV) per mission opportunity - a very Technology development cost of the nuclear reactor for an NEP system would be shared with the proposed nuclear surface power systems, since nuclear systems will be required to provide substantial electrical power on the surface of Mars. The NEP development project plan proposed includes evolutionary technology development for nuclear electric propulsion systems that expands upon SP-100 (Space Power - 100 kw(e)) technology that has been developed for lunar and Mars surface nuclear power, and small NEP systems for interplanetary probes. System upgrades are expected to evolve that will result in even shorter trip times, improved payload capabilities, and enhanced safety and reliability.

Clark, John S.↗

DOE OSTI2024

Vertical instability forecasting and controllability assessment of multi-device tokamak plasmas in DECAF with data-driven optimization

Abstract Reliable vertical position control will be an essential element of any future tokamak-based fusion power plant in order to reduce disruptions and maximize performance. We investigate methods to improve vertical controllability boundary determination in plasma operational space and demonstrate a data-driven approach based on direct pseudoinversion of operational space data that is rigorously quantitative, applicable in real-time plasma control systems, and physically intuitive to interpret. Applied to historical shot data from entire run campaigns on the MAST-U, KSTAR, and NSTX tokamaks, this approach, implemented in DECAF, improves vertical displacement event identification accuracy to 98.9%–100%. Further, we explore the application of a physics-based vertical stability metric as an early warning forecaster for vertical displacement events. The development of a linear surrogate model for the plasma current density profile, with a coefficient of determination of 0.992 on the training dataset, enables potential employment of this forecaster in real-time. The application of this approach on historical data from the MAST-U MU02 campaign yields a forecaster with 62.6% accuracy, indicating promise for this method when further refined and potentially coupled with other stability metrics.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

DOE OSTI2026

Evaluating multistation phase picking algorithm phase neural operator (PhaseNO) on local seismic networks

Reliable automatic phase picking is important for many seismic applications. With the development of machine learning approaches, many algorithms are proposed, evaluated and applied to different areas. Many of these algorithms are single station based, while recent proposed methods start to combine surrounding stations into consideration in the problem of phase picking. Among these algorithms, the phase neural operator (PhaseNO) shows promising results on regional data sets comparing to existing algorithms. But there are many use cases for the local seismic networks in our community, therefore in this paper we evaluate the performance of PhaseNO on four different local data sets and compare the results to PhaseNet and EQTransformer. We used both individual phase picking metrics as well as association metrics to illustrate the performance of PhaseNO. By manually reviewing the newly detected events, we find that the PhaseNO model outperforms the single station-based approaches in the local-scale use cases due to its consideration of coherent signals from multiple stations. We also explored PhaseNO’s behaviours when only using one station, as well as gradually increasing the number of stations in the seismic network to better understand its behaviour. Overall, using the off-the-shelf machine learning based phase pickers, PhaseNO demonstrated its good performance on local-scale seismic networks.

58 GEOSCIENCES↗

DOE OSTI2021

FPV Video Adaptation for UAV Collision Avoidance

First person view (FPV) technology for unmanned aerial vehicles (UAVs) provides an immersive experience for pilots and enables various personal and commercial applications such as aerial photography, drone racing, search and rescue operations, agricultural surveillance, and structural inspection. While real time video streaming from a UAV and vision-based collision avoidance strategies have been studied in literature as separate topics, in this paper we tackle collision avoidance in FPV scenarios, taking into account network delays and real time video parameters. We present a theoretical model for obstacle collisions that considers the current communication channel conditions, the real time video parameters, and the UAV's position relative to the closest obstacle. A video adaptation algorithm is then designed, using this metric, to tune the FPV video resolution, number of re-transmission attempts, and the modulation scheme to maximize the probability of avoiding collisions. This algorithm also takes into account specific latency constraints of the application. This video algorithm was evaluated in various scenarios and its ability to respond to both distances to the obstacle as well as the communication channel conditions was demonstrated. It was found that, for the considered scenarios, the performance of the proposed adaptive algorithm was, on an average, 58.63% higher than the closest non-adaptive one in terms of maximizing the probability of avoiding collision. Such collision avoidance strategies could be used to make UAV FPV applications safer and more reliable.

47 OTHER INSTRUMENTATION↗

DOE OSTI2026

Deployment of Traditional and Hybrid Machine Learning for Critical Heat Flux Prediction in the CTF Thermal-Hydraulics Code

Critical heat flux (CHF) marks the transition from nucleate to film boiling, where heat transfer to the working fluid can rapidly deteriorate. Accurate CHF prediction is essential for efficiency, safety, and preventing equipment damage, particularly in nuclear reactors. Although widely used, empirical correlations frequently exhibit discrepancies when compared to experimental data, limiting their reliability in diverse operational conditions. Traditional machine learning (ML) approaches have demonstrated potential for CHF prediction but often suffer from limited interpretability, data scarcity, and insufficient knowledge of physical principles. Hybrid model approaches, which combine data-driven ML with base models, mitigate these concerns by incorporating prior knowledge of the domain. This study integrates an externally trained purely data-driven ML model and two hybrid models (using the Biasi and Bowring CHF correlations) within the CTF subchannel code via a custom Fortran framework. Performance was evaluated using two validation cases: a subset of the Nuclear Regulatory Commission (NRC) CHF database and the Bennett dryout experiments. In both cases, the hybrid models demonstrated significantly lower error metrics compared to conventional empirical correlations, with the best models often reducing relative error by about 5 percentage points. The pure ML model achieved comparable accuracy, outperforming the hybrid Biasi model in the NRC test case (3.3% versus 5.5% relative error) but exhibiting slightly higher error against the hybrid Bowring model in the Bennett test case (7.7% versus 6.1%). Trend analysis of error parity indicated that ML-based models reduced the tendency for CHF overprediction, improving overall accuracy. These results demonstrate that ML-based CHF models can be effectively integrated into subchannel codes and could potentially increase performance compared to conventional methods.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗

DOE OSTI2026

Machine Learning for Mapping Multipactor Susceptibility in RF Systems: Capabilities and Generalization Constraints

Multipactor is a surface-driven electron avalanche phenomenon that degrades the performance and reliability of radio-frequency (RF) systems in particle accelerator and vacuum electronics applications. Multipactor behavior in a given device structure is conventionally assessed through susceptibility charts, which provide a parameter-space characterization of the instability. In this work, we assess the capabilities of machine-learning (ML) models to learn and predict such susceptibility charts and analyze the constraints governing their generalization across materials. Using a simulation-derived dataset spanning six distinct secondary-electron-yield material profiles in a canonical two-surface planar geometry, we train supervised regression models and artificial neural networks to predict the time-averaged electron growth rate, δavg, across the relevant parameter space. Model performance is evaluated using metrics that explicitly probe the structure of susceptibility charts, including Intersection over Union, Structural Similarity Index, and correlation analysis. Tree-based ensemble models outperform neural-network models in reconstructing susceptibility regions and in generalizing across material domains. Principal-component analysis reveals disjoint material feature distributions, indicating that the piecewise mode structure of multipactor susceptibility is difficult to represent with a single global model and that generalization is constrained by data coverage rather than by model complexity. An exhaustive reduced-coverage study further shows that sparse material-space coverage can yield mean performance in the same general range but producing large variability in the susceptibility-region overlap. These results clarify the capabilities of ML-based surrogate models for parameter-space characterization of multipactor discharge. They also provide guidance for their appropriate use in RF system design.

43 PARTICLE ACCELERATORS↗