Search NASA⌕ Search

SEARCH · Search NASA

Results for “Operator 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 181 records · Page 10
All sourcesNASA NTRSDOE OSTI
DOE OSTI2023

Quantifying Uncertainty of Deep Reinforcement Learning Based Decision Making for Operations and Maintenance of Nuclear Power Plant

This paper summarizes research that integrates condition monitoring and prognostics with decision making for nuclear power plant operations and maintenance. As part of this research, we have developed an online asset management tool to help reduce life-cycle maintenance and repair costs. Using the latest advancements in condition monitoring, supply chain analytics, and deep reinforcement learning, we have created a predictive maintenance tool that can optimize the maintenance and spare-part management of a repairable nuclear system. To demonstrate these methods, preliminary studies were conducted on a simple, representative maintenance system undergoing a stochastic degradation process that requires repairs or replacement to continue operation. Through Monte Carlo simulations, we were able to reduce maintenance spending by approximately 50% compared to optimized, time-based maintenance strategies. Not only does the decision maker reduce the average life-cycle costs, it also minimizes the chance of high cost scenarios, lowering the variance of the expected cost distributions, and reducing overall financial risk. Furthermore, this work also studies the ability of the decision maker to handle various levels of noise from observation uncertainty. By introducing uncertainty into the decision-making process, we have quantified the robustness and resiliency of the decision maker, as well as identified necessary levels of observability to demonstrate cost effectiveness.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗
DOE OSTI2022

Physiochemical Machine Learning Models Predict Operational Lifetimes of CH3NH3PbI3 Perovskite Solar Cells

Halide perovskites are promising photovoltaic (PV) materials with the potential to lower the cost of electricity and greatly expand the penetration of PV if they can demonstrate long-term stability under illumination in the presence of moisture and oxygen. The solar cell service lifetime as quantified by the T80 (the time required for the power conversion efficiency to drop to 80% of its starting value) is a useful metric to assess stability. The T80 for utility, commercial, or residential PV systems needs to be several decades in order to yield low-cost electricity, and thus it is not practical to directly measure the T80. It would be useful if T80 could be predicted from the initial dynamics of a solar cell’s performance, but until now no models have been developed to forecast T80. In this work, we report the development of machine learning models to predict T80 of ITO/NiOx/CH3NH3PbI3/C60/BCP/Ag solar cells operating at maximum power point under 1-sun equivalent photon flux in air at varying temperatures and relative humidities. Efficiency losses are driven by short-circuit current and fill factor, indicating that chemical decomposition of the perovskite is a major contributor to degradation. Spatial patterns evident from in situ dark field optical microscopy suggest that the electric field gradient at device edges plays a significant role in perovskite decomposition, along with photochemical reactions with O2 and H2O. Models are trained using a menu of features from three distinct categories: (i) features based on measurements of the initial rates of change of device parameters, (ii) features based on the ambient conditions during operation (temperature, & partial pressure of H2O), and (iii) features based on underlying physics and chemistry. We show that a theory-based physiochemical feature derived from a model of the chemical reaction kinetics of the rate of degradation of the CH3NH3PbI3 is particularly valuable for prediction. This physiochemical feature was selected as the first or second most dominant feature in the best performing models. With a dataset consisting of 45 accelerated degradation experiments with T80 that range over a factor of 30, the model predicts T80 with an accuracy of about 40% (|predicted T80 - observed T80| / observed T80) on samples not used in training. This hybrid ML approach should be effective when applied to other compositions, device architectures, and advanced packaging schemes.

14 SOLAR ENERGY↗
DOE OSTI2022

Physiochemical Machine Learning Models Predict Operational Lifetimes of CH3NH3PbI3 Perovskite Solar Cells

Halide perovskites are promising photovoltaic (PV) materials with the potential to lower the cost of electricity and greatly expand the penetration of PV if they can demonstrate long-term stability under illumination in the presence of moisture and oxygen. The solar cell service lifetime as quantified by the T80 (the time required for the power conversion efficiency to drop to 80% of its starting value) is a useful metric to assess stability. The T80 for utility, commercial, or residential PV systems needs to be several decades in order to yield low-cost electricity, and thus it is not practical to directly measure the T80. It would be useful if T80 could be predicted from the initial dynamics of a solar cell’s performance, but until now no models have been developed to forecast T80. In this work, we report the development of machine learning models to predict T80 of ITO/NiOx/CH3NH3PbI3/C60/BCP/Ag solar cells operating at maximum power point under 1-sun equivalent photon flux in air at varying temperatures and relative humidities. Efficiency losses are driven by short-circuit current and fill factor, indicating that chemical decomposition of the perovskite is a major contributor to degradation. Spatial patterns evident from in situ dark field optical microscopy suggest that the electric field gradient at device edges plays a significant role in perovskite decomposition, along with photochemical reactions with O2 and H2O. Models are trained using a menu of features from three distinct categories: (i) features based on measurements of the initial rates of change of device parameters, (ii) features based on the ambient conditions during operation (temperature, & partial pressure of H2O), and (iii) features based on underlying physics and chemistry. We show that a theory-based physiochemical feature derived from a model of the chemical reaction kinetics of the rate of degradation of the CH3NH3PbI3 is particularly valuable for prediction. This physiochemical feature was selected as the first or second most dominant feature in the best performing models. With a dataset consisting of 45 accelerated degradation experiments with T80 that range over a factor of 30, the model predicts T80 with an accuracy of about 40% (|predicted T80 - observed T80| / observed T80) on samples not used in training. This hybrid ML approach should be effective when applied to other compositions, device architectures, and advanced packaging schemes.

14 SOLAR ENERGY↗
NASA NTRS2006

The International Space Station: Operations and Assembly - Learning From Experiences - Past, Present, and Future

As the Space Shuttle continues flight, construction and assembly of the International Space Station (ISS) carries on as the United States and our International Partners resume the building, and continue to carry on the daily operations, of this impressive and historical Earth-orbiting research facility. In his January 14, 2004, speech announcing a new vision for America s space program, President Bush ratified the United States commitment to completing construction of the ISS by 2010. Since the launch and joining of the first two elements in 1998, the ISS and the partnership have experienced and overcome many challenges to assembly and operations, along with accomplishing many impressive achievements and historical firsts. These experiences and achievements over time have shaped our strategy, planning, and expectations. The continual operation and assembly of ISS leads to new knowledge about the design, development and operation of systems and hardware that will be utilized in the development of new deep-space vehicles needed to fulfill the Vision for Exploration and to generate the data and information that will enable our programs to return to the Moon and continue on to Mars. This paper will provide an overview of the complexity of the ISS Program, including a historical review of the major assembly events and operational milestones of the program, along with the upcoming assembly plans and scheduled missions of the space shuttle flights and ISS Assembly sequence.

Fuller, Sean↗
NASA NTRS2005

X-43A Fluid and Environmental Systems: Ground and Flight Operation and Lessons Learned

The X-43A Hyper-X program demonstrated the first successful flights of an airframe integrated scramjet powered hypersonic vehicle. The X-43A vehicles established successive world records for jet-powered vehicles at speeds of Mach 7 and Mach 10. The X-43A vehicle is a subscale version of proposed hypersonic reconnaissance strike aircraft. Scaled down to a length of 12 ft (3.66 m), the lifting body design with high fineness ratio resulted in very small internal space available for fluid systems and their corresponding environmental conditioning systems. Safe testing and operation of the X-43A fluid and environmental systems was critical for mission success, not only for the safety of the flight crew in the NASA B-52B carrier aircraft, but also to maintain the reliability of vehicle systems while exposed to dynamics and hostile conditions encountered during the boost trajectory. The X-43A fluid and environmental systems successfully managed explosive, pyrophoric, inert, and very high pressure gases without incident. This report presents a summary of the checkout and flight validation of the X-43A fluid systems. The testing used for mission assurance is summarized. System performance during captive carry and launch flights is presented. The lessons learned are also discussed.

Vachon, Michael Jacob↗
NASA NTRSDate not supplied

Lessons Learned from Astrobee Operations on the International Space Station

Since its launch in 2019, NASA has been operating three Astrobee free-flying robots providing an autonomous and adaptable research platform aboard the International Space Station (ISS). These robots have not only facilitated a myriad of national and international research endeavors in microgravity but have also served as a STEM outreach platform for student competitions aboard the ISS. Amidst its extensive operational tenure, spanning over five years and exceeding 1200 hours of cumulative free-flyer operation as of April 2024, the Astrobee robots have encountered software and hardware anomalies. Despite its inherent design for on-orbit repair or replacement, certain anomalies have proven to be complex, necessitating remote resolution via software and firmware updates or, in extreme cases, hardware replacements or the return of faulty units to NASA's ground facilities for repair. Such challenges underscore the delicate balance between the autonomous functionality of Astrobee and the occasional need for human intervention to maintain optimal performance. One recurring point of failure identified during Astrobee's operational lifespan has been the SD card, a critical component utilized by the different Astrobee processors and the Dock Station. The occurrence of SD card anomalies, both on orbit and within ground units, has provided invaluable insights into the improvement of Astrobee's systems and mitigation to future faults. This presentation will focus on four key areas: 1. Overview of Faults and Anomalies: A comprehensive examination of the diverse array of faults and anomalies encountered by Astrobee and its associated systems both in orbit and on the ground. From software glitches to hardware malfunctions, this section provides insights into the challenges faced during Astrobee's operational tenure. 2. Resolution Processes and Procedures: An in-depth discussion of the methodologies and procedures implemented to resolve the encountered anomalies. This includes remote troubleshooting, software patches, firmware updates, and, when necessary, the logistics involved in hardware replacements or down-massing for repair. 3. Implementation of Software Updates and Hardware Upgrades: A detailed exploration of the strategies employed to mitigate the risk of recurring anomalies through the implementation of software updates and hardware upgrades. This section highlights the iterative nature of Astrobee's development, emphasizing the continuous pursuit of robustness and reliability. 4. Lessons Learned and Future Directions: Reflecting on the insights gained from addressing anomalies, this section examines the lessons learned and outlines future directions for enhancing Astrobee's robustness and resilience. It underscores the iterative nature of space exploration and the importance of adaptability and continuous improvement in the pursuit of scientific discovery. Through a nuanced examination of Astrobee's operational challenges and the strategies employed to overcome them, this presentation sheds light on the complexities of operating autonomous robotic systems in the ISS environment. It underscores NASA's commitment to pushing the boundaries of exploration and innovation while navigating the inherent challenges of space exploration.

Astrobee↗
NASA NTRSDate not supplied

Lessons Learned from Astrobee Operations on the International Space Station

Since its launch in 2019, NASA has been operating three Astrobee free-flying robots providing an autonomous and adaptable research platform aboard the International Space Station (ISS). These robots have not only facilitated a myriad of national and international research endeavors in microgravity but have also served as a STEM outreach platform for student competitions aboard the ISS. Amidst its extensive operational tenure, spanning over five years and exceeding 1200 hours of cumulative free-flyer operation as of April 2024, the Astrobee robots have encountered software and hardware anomalies. Despite its inherent design for on-orbit repair or replacement, certain anomalies have proven to be complex, necessitating remote resolution via software and firmware updates or, in extreme cases, hardware replacements or the return of faulty units to NASA's ground facilities for repair. Such challenges underscore the delicate balance between the autonomous functionality of Astrobee and the occasional need for human intervention to maintain optimal performance. One recurring point of failure identified during Astrobee's operational lifespan has been the SD card, a critical component utilized by the different Astrobee processors and the Dock Station. The occurrence of SD card anomalies, both on orbit and within ground units, has provided invaluable insights into the improvement of Astrobee's systems and mitigation to future faults. This presentation will focus on four key areas: 1. Overview of Faults and Anomalies: A comprehensive examination of the diverse array of faults and anomalies encountered by Astrobee and its associated systems both in orbit and on the ground. From software glitches to hardware malfunctions, this section provides insights into the challenges faced during Astrobee's operational tenure. 2. Resolution Processes and Procedures: An in-depth discussion of the methodologies and procedures implemented to resolve the encountered anomalies. This includes remote troubleshooting, software patches, firmware updates, and, when necessary, the logistics involved in hardware replacements or down-massing for repair. 3. Implementation of Software Updates and Hardware Upgrades: A detailed exploration of the strategies employed to mitigate the risk of recurring anomalies through the implementation of software updates and hardware upgrades. This section highlights the iterative nature of Astrobee's development, emphasizing the continuous pursuit of robustness and reliability. 4. Lessons Learned and Future Directions: Reflecting on the insights gained from addressing anomalies, this section examines the lessons learned and outlines future directions for enhancing Astrobee's robustness and resilience. It underscores the iterative nature of space exploration and the importance of adaptability and continuous improvement in the pursuit of scientific discovery. Through a nuanced examination of Astrobee's operational challenges and the strategies employed to overcome them, this presentation sheds light on the complexities of operating autonomous robotic systems in the ISS environment. It underscores NASA's commitment to pushing the boundaries of exploration and innovation while navigating the inherent challenges of space exploration.

Astrobee↗
NASA NTRS2017

Lessons Learned in Science Operations for Planetary Surface Exploration

The six Apollo lunar surface missions represent the only occasions where we have conducted scientific operations on another planetary surface. While these six missions were successful in bringing back valuable geologic samples, technology advances in the subsequent forty years have enabled much higher resolution scientific activity in situ. Regardless of where astronauts next visit (whether it be back to the Moon or to Mars or a Near Earth Object), the science operations procedures completed during this mission will need to be refined and updated to reflect these advances. We have undertaken a series of operational tests in relevant field environments to understand how best to develop the new generation of science operations procedures for planetary surface exploration.

Young, K. E.↗
NASA NTRS2019

Using Machine-Learning to Dynamically Generate Operationally Acceptable Strategic Reroute Options

The newly developed Trajectory Option Set (TOS), a preference-weighted set of alternative routes submitted by flight operators, is a capability in the U.S. traffic flow management system that enables automated trajectory negotiation between flight operators and Air Navigation Service Providers. The objective of this paper is to describe and demonstrate an approach for automatically generating pre-departure and airborne TOSs that have a high probability of operational acceptance. The approach uses hierarchical clustering of historical route data to identify route candidates. The probability of operational acceptance is then estimated using predictors trained on historical flight plan amendment data using supervised machine learning algorithms, allowing the routes with highest probability of operational acceptance to be selected for the TOS. Features used describe historical route usage, difference in flight time and downstream demand to capacity imbalance. A random forest was found to be the best performing algorithm for learning operational acceptability, with a model accuracy of 0.96. The approach is demonstrated for an historical pre-departure flight from Dallas/Fort Worth International Airport to Newark Liberty International Airport.

Evans, Antony↗
NASA NTRS2019

Using Machine-Learning to Dynamically Generate Operationally Acceptable Strategic Reroute Options

The newly developed Trajectory Option Set (TOS), a preference-weighted set of alternative routes submitted by flight operators, is a capability in the U.S. traffic flow management system that enables automated trajectory negotiation between flight operators and Air Navigation Service Providers. The objective of this paper is to describe and demonstrate an approach for automatically generating pre-departure and airborne TOSs that have a high probability of operational acceptance. The approach uses hierarchical clustering of historical route data to identify route candidates. The probability of operational acceptance is then estimated using predictors trained on historical flight plan amendment data using supervised machine learning algorithms, allowing the routes with highest probability of operational acceptance to be selected for the TOS. Features used describe historical route usage, difference in flight time and downstream demand to capacity imbalance. A random forest was found to be the best performing algorithm for learning operational acceptability, with a model accuracy of 0.96. The approach is demonstrated for an historical pre-departure flight from Dallas/Fort Worth International Airport to Newark Liberty International Airport.

Evans, Antony↗