Search NASA⌕ Search

SEARCH · Search NASA

Results for “Adaptive learning”

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 19 records
All sourcesNASA NTRSDOE OSTI
DOE OSTI2026

A score-based diffusion model approach for adaptive learning of stochastic partial differential equation solutions

In this paper, we propose a novel framework for adaptively learning the time-evolving solutions of stochastic partial differential equations (SPDEs) using score-based diffusion models within a recursive Bayesian inference setting. SPDEs play a central role in modeling complex physical systems under uncertainty, but their numerical solutions often suffer from model errors and reduced accuracy due to incomplete physical knowledge and environmental variability. To address these challenges, we encode the governing physics into the score function of a diffusion model using simulation data and incorporate observational information via a likelihood-based correction in a reverse-time stochastic differential equation. This enables adaptive learning through iterative refinement of the solution as new data becomes available. To improve computational efficiency in high-dimensional settings, we introduce the ensemble score filter, a training-free approximation of the score function designed for real-time inference. Numerical experiments on benchmark SPDEs demonstrate the accuracy and robustness of the proposed method under sparse and noisy observations.

97 MATHEMATICS AND COMPUTING↗
NASA NTRS1988

Strategies for adding adaptive learning mechanisms to rule-based diagnostic expert systems

Rule-based diagnostic expert systems can be used to perform many of the diagnostic chores necessary in today's complex space systems. These expert systems typically take a set of symptoms as input and produce diagnostic advice as output. The primary objective of such expert systems is to provide accurate and comprehensive advice which can be used to help return the space system in question to nominal operation. The development and maintenance of diagnostic expert systems is time and labor intensive since the services of both knowledge engineer(s) and domain expert(s) are required. The use of adaptive learning mechanisms to increment evaluate and refine rules promises to reduce both time and labor costs associated with such systems. This paper describes the basic adaptive learning mechanisms of strengthening, weakening, generalization, discrimination, and discovery. Next basic strategies are discussed for adding these learning mechanisms to rule-based diagnostic expert systems. These strategies support the incremental evaluation and refinement of rules in the knowledge base by comparing the set of advice given by the expert system (A) with the correct diagnosis (C). Techniques are described for selecting those rules in the in the knowledge base which should participate in adaptive learning. The strategies presented may be used with a wide variety of learning algorithms. Further, these strategies are applicable to a large number of rule-based diagnostic expert systems. They may be used to provide either immediate or deferred updating of the knowledge base.

Stclair, D. C.↗
NASA NTRSDate not supplied

Adaptive Learning for Reliability Analysis using Support Vector Machines

A novel algorithm is presented for adaptive learning of an unknown function that separates two regions of a domain.In the context of reliability analysis these two regions represent the failure domain, where a set of constraints or requirements are violated, and a safe domain where they are satisfied. The Limit State Function (LSF) separates these two regions. Evaluating the constraints for a given parameter point requires the evaluation of a computational model that may well be expensive. For this reason we wish to construct a meta-model that can estimate the LSFas accurately as possible, using only a limited amount of training data. This work presents an adaptive strategy employing a Support Vector Machine (SVM) as a meta-model to provide a semi-algebraic approximation of the LSF.We describe an optimization process that is used to select informative parameter points to add to training data at each iteration to improve the accuracy of this approximation. A formulation is introduced for bounding the predictions of the meta-model; in this way we seek to incorporate this aspect of Gaussian Process Models (GPMs) within anSVM meta-model. Finally, we apply our algorithm to two benchmark test cases, demonstrating performance that is comparable with, if not superior, to a standard technique for reliability analysis that employs GPMs

Adaptive learning↗
NASA NTRS1976

An adaptive learning control system for aircraft

A learning control system is developed which blends the gain scheduling and adaptive control into a single learning system that has the advantages of both. An important feature of the developed learning control system is its capability to adjust the gain schedule in a prescribed manner to account for changing aircraft operating characteristics. Furthermore, if tests performed by the criteria of the learning system preclude any possible change in the gain schedule, then the overall system becomes an ordinary gain scheduling system. Examples are discussed.

Mekel, R.↗
NASA NTRS1985

An adaptive learning control system for large flexible structures

The objective of the research has been to study the design of adaptive/learning control systems for the control of large flexible structures. In the first activity an adaptive/learning control methodology for flexible space structures was investigated. The approach was based on using a modal model of the flexible structure dynamics and an output-error identification scheme to identify modal parameters. In the second activity, a least-squares identification scheme was proposed for estimating both modal parameters and modal-to-actuator and modal-to-sensor shape functions. The technique was applied to experimental data obtained from the NASA Langley beam experiment. In the third activity, a separable nonlinear least-squares approach was developed for estimating the number of excited modes, shape functions, modal parameters, and modal amplitude and velocity time functions for a flexible structure. In the final research activity, a dual-adaptive control strategy was developed for regulating the modal dynamics and identifying modal parameters of a flexible structure. A min-max approach was used for finding an input to provide modal parameter identification while not exceeding reasonable bounds on modal displacement.

Thau, F. E.↗
DOE OSTI2026

Learned adaptive properties for mitigation of weight perturbations in embedded spiking networks

Recent years have seen an increased importance of neural network inference in edge-based scenarios, which impose size and power constraints requiring novel computing devices. These same edge scenarios may require operating over long periods of time, or exposure to extreme environments, resulting in a drift of neural network weights that cause degraded performance. In searching for ways to develop neural network approaches that perform robustly under these conditions, we propose a biologically-inspired mechanism for the dynamic adaptation of within-neuron parameters that is guided by a global context signal carrying information about perturbations and variability in incoming stimuli. Specifically, we demonstrate that adaptive voltage thresholds or neuronal time constants, when informed by a global context signal, can enable network-level mechanisms to recover from perturbed synaptic weights. Consistent with prior literature, the context-modulated approach is effective for recurrent, but not feedforward networks, by modulating network level dynamics. We demonstrate this approach successfully recovers performance in image classification tasks and spatiotemporal tracking tasks under idealized and Gaussian noise as well as for realistic perturbations from a memristive device when exposed to ionizing radiation. Finally, we discuss how this approach enables the design of robust and energy-efficient neuromorphic systems that perform well, even in resource-constrained scenarios with extreme environments such as edge processing.

context modulation↗
NASA NTRS1978

An adaptive learning control system for aircraft

A learning control system and its utilization as a flight control system for F-8 Digital Fly-By-Wire (DFBW) research aircraft is studied. The system has the ability to adjust a gain schedule to account for changing plant characteristics and to improve its performance and the plant's performance in the course of its own operation. Three subsystems are detailed: (1) the information acquisition subsystem which identifies the plant's parameters at a given operating condition; (2) the learning algorithm subsystem which relates the identified parameters to predetermined analytical expressions describing the behavior of the parameters over a range of operating conditions; and (3) the memory and control process subsystem which consists of the collection of updated coefficients (memory) and the derived control laws. Simulation experiments indicate that the learning control system is effective in compensating for parameter variations caused by changes in flight conditions.

Mekel, R.↗
NASA NTRS1979

Aircraft adaptive learning control

The optimal control theory of stochastic linear systems is discussed in terms of the advantages of distributed-control systems, and the control of randomly-sampled systems. An optimal solution to longitudinal control is derived and applied to the F-8 DFBW aircraft. A randomly-sampled linear process model with additive process and noise is developed.

Lee, P. S. T.↗
NASA NTRS1986

An application of adaptive learning to malfunction recovery

A self-organizing controller is developed for a simplified two-dimensional aircraft model. The Controller learns how to pilot the aircraft through a navigational mission without exceeding pre-established position and velocity limits. The controller pilots the aircraft by activating one of eight directional actuators at all times. By continually monitoring the aircraft's position and velocity with respect to the mission, the controller progressively modifies its decision rules to improve the aircraft's performance. When the controller has learned how to pilot the aircraft, two actuators fail permanently. Despite this malfunction, the controller regains proficiency at its original task. The experimental results reported show the controller's capabilities for self-organizing control, learning, and malfunction recovery.

Cruz, R. E.↗
NASA NTRS1980

Adaptive and learning control of large space structures

The paper describes the adaptive learning system for space operations which assumes that structural testing can be conducted during deployment and assembly. Simulation results using the solar electric propulsion array and a novel remote sensor are presented; they involve faster scan television coverage of the motions of the array from four cameras on the corners of the Space Shuttle payload bay. The description of the simulation, the filtering algorithm for processing the TV data, the parameter extraction algorithm, and the simulation results are presented.

Montgomery, R. C.↗
DOE OSTI2026

Adaptive Reinforcement Learning Control for Power Distribution in Multi-Output Resonant Converters

This paper presents an adaptive reinforcement learning (ARL)-based control framework for efficient power distribution in a multi-output resonant converter for UAV applications. The proposed system is based on a high-frequency isolated resonant architecture, where a single energy source supplies multiple propulsion loads through independently controlled output rectifiers, addressing the need for coordinated multi-motor power management. The ARL framework dynamically allocates output power by learning optimal phase-shift control actions under varying load demands and operating conditions. The agent autonomously determines control parameters that maximize conversion efficiency while ensuring accurate power sharing among multiple outputs. In addition, the proposed approach enables adaptive operation without requiring detailed system modeling or manual tuning. Experimental results demonstrate stable and efficient performance over a wide range of operating conditions, confirming the effectiveness and robustness of the learning-based control strategy for multi-output resonant converter system.

Asa, Erdem [ORNL] (ORCID:0000000190884812)↗
DOE OSTI2026

Adaptive Reinforcement Learning (ARL) Control of a Multi-port Resonant Converter in UAV Systems

This study presents an adaptive reinforcement learning (ARL) control framework for a multi-port resonant converter used in hybrid unmanned aerial vehicle (UAV) power systems. The converter integrates high-frequency half-bridge input ports connected to a rectified engine–generator set and a battery energy storage system, along with a semi-bridgeless active rectifier supplying the propulsion load. A deep RL agent is trained to dynamically regulate inter-port phase-shift commands in real time based on flight conditions and load power demand. The ARL controller autonomously identifies phase-shift combinations that maximize conversion efficiency while maintaining stable and coordinated power flow, even under rapidly varying operating scenarios. This data-driven approach eliminates the need for explicit system modeling or extensive manual tuning and enables coordinated control among multiple power ports without inter-port communication. Experimental results validate that the ARL based strategy achieves reliable power sharing and consistently high-efficiency operation across diverse UAV operating conditions.

Asa, Erdem [ORNL] (ORCID:0000000190884812)↗
NASA NTRSDate not supplied

Machine-Learning-Based Adaptive Thinning of CrIS Radiances to Improve Global Tropical Cyclone Analysis and Forecasts

This work is focused on optimizing the assimilation of hyperspectral infrared (IR) radiances from the Cross-track Infrared Sounder (CrIS) with the goal of improving the representation of tropical cyclones (TCs) in global analyses and forecasts. Current operational assimilation systems rely on subsampling IR radiances on a regular thinning grid. A new and improved adaptive methodology based on machine learning (ML) recognizes TCs from geostationary satellite imagery and is implemented in the Goddard Earth Observing System (GEOS) model and data assimilation framework. The ML methodology is extensively trained on existing TC data sets and creates for each TC a dynamic mask, based on the evolving shape and life cycle of that specific event. Once a TC mask is created, a switch is then activated in the data assimilation system to alter the thinning, ingesting more CrIS radiances within the moving mask, thus increasing the TC sampling. After the TC dissipates, the assimilation of CrIS radiances reverts to normal data density. Results of TC segmentation provided by a state-of-the-art generative machine learning model known as the Denoising Diffusion Probabilistic Model (DDPM) are compared to the previously used U-Net model. The new approach surpasses the performance of the previously developed one. The methodology is applied to both clear-sky and cloud-cleared radiances. Benefits from the latter methodology, particularly in improving the structure of TCs and the intensity forecasts, are presented.

Oreste Reale↗
NASA NTRSDate not supplied

Adaptive Pressure Profile Method to Locate the Isolator Shock Train Leading Edge Given Limited Pressure Information

To maximize the performance of high-speed air-breathing engines, such as dual-mode scramjets, the streamwise location of the shock train leading edge (STLE) is ideally placed as far upstream in the isolator as possible while avoiding engine unstart. Thus, it is of interest to quantify and control the STLE location as the vehicle travels along its flight trajectory. The STLE location is typically quantified using wall static pressure measurements but there are often restrictions on the number and placement of transducers, thus reducing the accuracy and overall capability of STLE detection methods. In this work, the Adaptive Pressure Profile (APP) method is introduced to address such sparsity concerns. This method is data driven and does not heavily rely on prior information about the flow regime or engine model. Instead, the \mname method uses real-time pressure measurements from a small number of transducers to adaptively learn the isolator pressure profile. This adaptively-learned profile is fit to the pressure data at each time instance to estimate STLE location. The \mname method produces accurate estimates even when (1) the STLE location is not bounded by two or more transducers or (2) when the STLE location is between two transducers that are situated several duct heights apart. Data from two direct-connect isolator models are used to evaluate the accuracy of the \mname method and demonstrate its robustness for different back-pressure scenarios and transducer configurations.

Robin L Hunt↗
NASA NTRSDate not supplied

Adaptive Pressure Profile Method to Locate the Isolator Shock Train Leading Edge Given Limited Pressure Information

To maximize the performance of high-speed air-breathing engines, such as dual-mode scramjets, the streamwise location of the shock train leading edge (STLE) is ideally placed as far upstream in the isolator as possible while avoiding engine unstart. Thus, it is of interest to quantify and control the STLE location as the vehicle travels along its flight trajectory. The STLE location is typically quantified using wall static pressure measurements but there are often restrictions on the number and placement of transducers, thus reducing the accuracy and overall capability of STLE detection methods. In this work, the Adaptive Pressure Profile (APP) method is introduced to address such sparsity concerns. This method is data driven and does not heavily rely on prior information about the flow regime or engine model. Instead, the \mname method uses real-time pressure measurements from a small number of transducers to adaptively learn the isolator pressure profile. This adaptively-learned profile is fit to the pressure data at each time instance to estimate STLE location. The \mname method produces accurate estimates even when (1) the STLE location is not bounded by two or more transducers or (2) when the STLE location is between two transducers that are situated several duct heights apart. Data from two direct-connect isolator models are used to evaluate the accuracy of the \mname method and demonstrate its robustness for different back-pressure scenarios and transducer configurations.

Robin Hunt↗