100039 research outputs found
Sort by
Toward Trusted Onboard AI: Advancing Small Satellite Operations using Reinforcement Learning
No full textA Reinforcement Learning (RL) algorithm was developed for command automation onboard a 3U CubeSat. This effort focused on the implementation of macro Control Action Reinforcement Learning (CARL), a technique in which an onboard agent is provided with compiled information based on live telemetry as its observation. The agent uses this information to produce high-level actions, such as adjusting attitude to solar pointing, which are then translated into control algorithms and executed through lower-level instructions. Once trust in the onboard agent is established, real-time environmental information can be leveraged for faster response times and reduced reliance on ground control. The approach not only focuses on developing an RL algorithm for a specific satellite but also sets a precedent for integrating trusted artificial intelligence (AI) into onboard systems. This research builds on previous work in three areas: (1) RL algorithms for issuing high-level commands that are translated into low-level executable instructions; (2) the deployment of AI inference models interfaced with live operational systems, particularly onboard spacecraft; and (3) strategies for building trust in AI systems, especially for remote and autonomous applications. Existing RL research for satellite control is largely limited to simulation-based experiments; in this work, these techniques are tailored by constructing a digital twin of a specific spacecraft and training the RL agent to issue macro actions in this simulated environment. The policy of the trained agent is copied to an isolated environment, where it is fed compiled information about the satellite to make inference predictions, thereby demonstrating the RL algorithm’s validity on orbit without granting it command authority. This process enables safe comparison of the algorithm’s predictions against actual satellite behavior and ensures operation within expected parameters
Development of a Novel Thermal Management System for SmallSats
No full textIntroduction
- High-power space missions demand advanced thermal management
- Traditional rigid radiators limit scalability due to mass and volume.
- This work demonstrates a Thermal Management System (TMS) prototype integrating: Rollout Deployable Radiator (RDR) Phase Change Material Thermal Accumulator (PCM-TA) Integrated Heat Pipes (IHP)
- Components can be used individually or jointly
Exploring Beamed Energy Propulsion: Development and Application of a Configurable Simulation Tool
No full textLaser sailing enables propellantless spacecraft propulsion by transferring momentum from a directed energy beam to a reflective sail. Unlike traditional chemical or electric propulsion systems, laser sails decouple propulsion performance from onboard mass and stored energy, presenting an opportunity for lightweight, scalable spacecraft maneuvering. This work develops and applies a simulation framework in FreeFlyer to evaluate the performance of laser sailing across a range of orbital missions, with emphasis on low-mass systems, such as CubeSats. The simulation model incorporates several critical physical effects, including beam divergence, atmospheric attenuation, sail material reflectivity and emissivity, thermal loading, and realistic station visibility. The tool supports customizable mission geometries, sail properties, and ground or orbital laser configurations, allowing for comprehensive trade studies. Forces from both solar radiation pressure and laser illumination are modeled independently and combined when applicable. Three case studies are presented: (1) a low-Earth orbit altitude climb using both solar and laser propulsion, (2) a medium Earth orbit (MEO) to geostationary Earth orbit (GEO) transfer, and (3) Earth escape from a low-Earth orbit. Results demonstrate that orbital laser stations outperform ground-based stations by an order of magnitude in both time-to-target and energy efficiency due to uninterrupted visibility and the absence of atmospheric losses. In the Earth escape case, an orbital laser station enabled escape within 15 days using a 50 MW beam and a 10 kg payload. The MEO-to-GEO maneuver achieved an effective specific impulse over 9600 s, surpassing many electric propulsion systems. In the LEO climb experiment, even a modest 4-meter-radius sail enabled meaningful altitude increases, demonstrating feasibility for CubeSat-scale missions. Sensitivity analyses examine the effects of station latitude, sail material properties, and orbital inclination, revealing performance degradation when geometric alignment is suboptimal. Atmospheric losses are found to be the most significant constraint for ground-based systems, with beam efficiencies rarely exceeding 15%. The simulation also quantifies energy conversion losses and highlights how angular misalignment reduces net impulse delivery. While laser sailing offers a promising pathway for high-∆V, low-mass propulsion, implementation challenges remain. These include generating and sustaining megawatt-scale laser power, accurately tracking small sailcraft over large distances, and establishing orbital or surface-based infrastructure. Nevertheless, laser propulsion could enable new classes of space missions for small spacecraft, from low-cost constellation maintenance to interplanetary flybys, by offering reusable infrastructure and decoupling propulsion from payload mass. The modeling framework developed in this work provides a foundation for future mission design and optimization studies
Efficacy and Social Validity of a Modified Good Behavior Game
No full textTeachers regularly face challenging behavior from students in their classrooms. Research shows that the good behavior game (GBG) is a strategy teachers can use in their classrooms to improve student behavior. However, in practice, very few teachers choose to use this strategy in their classrooms. One reason why teachers may not use this intervention regularly is that they find some parts of the intervention to be difficult to implement in the long-term. The purpose of this study was to modify the procedures of the GBG to reduce implementation effort for teachers and assess if the intervention continues to be effective. In addition, the study sought to understand how teachers and students perceive the intervention. Results showed that the modified version of the intervention produced large improvements in students’ behavior in all four of the participating classrooms. In addition, teachers and students viewed the intervention as effective. Teachers, overall, viewed the intervention as feasible and easy to implement but also discussed some features of the intervention that can impact continued use of the intervention in their classrooms
Improving Seasonal Precipitation Forecasts in the Western United States Through Statistical Downscaling
No full textSeasonal precipitation forecasts are essential tools for water management, especially in drought-prone areas like the Western United States. While global forecasting systems can predict precipitation months to years in advance, these forecasts are typically produced at a coarse spatial resolution due to high computational costs – especially when multiple forecast iterations, or ensemble members, are included. This coarse resolution limits their ability to capture localized weather features, such as those shaped by the complex terrain of the Intermountain West. Here, analog statistical downscaling is demonstrated as an effective approach to enhance the spatial resolution of operational seasonal forecasts provided by the North American Multi-Model Ensemble within this challenging region. Additionally, we find that downscaling individual ensemble members – rather than downscaling the ensemble mean – results in greater forecast skill. These findings demonstrate a low-cost method to improve seasonal forecasts, providing a valuable framework for enhancing coarse resolution products through downscaling
Intra-Stand Movement of Balsam Woolly Adelgid (\u3ci\u3eAdelges piceae\u3c/i\u3e) in the Intermountain West
Get PDFBalsam woolly adelgid (BWA, Adelges piceae (Ratz.)) is an invasive insect of true firs (Abies spp.). BWA has spread from initial introductions on the west coast of North America in the 1920s through the Intermountain West and has recently invaded Utah. White fir (A. concolor) and subalpine fir (A. lasiocarpa) are the two main species of concern in this region; however, subalpine fir has been found to be more susceptible to attack and subsequent damage. While long-range dispersal mechanisms and patterns are of interest for spread across regions and for predicting new infestations, very little is known of BWA movement within a stand once it is present. We measured GPS locations of all trees within eight quarter-hectare BWA-infested subalpine fir stands in northern throughout Utah and southern Idaho and quantified BWA intensity on each tree at multiple points in time using a severity rating system. We measured tree- and stand-level attributes and quantified bole infestation levels twice per growing season for two seasons to detect change in infestation. We identified traits of individual subalpine fir individuals and stands that are associated with infestation status, change, and risk of new infestations. These models suggest that at a stand scale, individual tree attributes as well as larger scale neighborhood attributes affect the dispersal of BWA within subalpine fir stands. Our study suggests that current management recommendations should be followed and long-term monitoring should remain a priority
Habitat Selection Analysis of Mule Deer to Determine Efficacy of the Coyote Bounty Program in Utah
No full textMule deer populations have been declining over the past couple of decades in many parts of the western United States. Often times, predators are blamed for the decreases in deer populations. Because of this, wildlife managers implement programs focused on reducing the number of predators on the landscape in an attempt to increase deer populations. However, predator removal programs have had mixed results, thus further investigation is needed to determine if coyote removals by the public through a bounty program have a positive effect on mule deer populations. My study investigated whether coyote removals through the bounty program in Utah were occurring in areas that are important for mule deer fawns. My results show a spatial mismatch between areas where coyotes were removed by the bounty, and areas where mule deer are raising vulnerable fawns. This suggests that the bounty program may be ineffective at my study site. Further assessment is warranted to determine if the Utah Bounty program is ineffective statewide
An Ant-Colony Approach to Scheduling Charging Sessions of Electric Vehicle Fleets With Heterogeneous Scheduling Constraints
Get PDFElectric vehicles and electric-vehicle fleets are gaining widespread usage. A major challenge of managing an electric-vehicle fleet is scheduling when the vehicles can charge and when the vehicles can complete their tasks. Finding smart ways to schedule electric-vehicle charging reduces both the cost of charging vehicles and increases the health of the grid and the environment. This best-of-both-worlds outcome is because of the pricing structure that electric utility companies use. When minimizing the cost of charging the vehicles, it is important to consider that vehicle fleets have constraints on when they can charge. Vehicles are usually limited by the tasks they need to perform. Examples of these tasks include buses attending to their stops and packages being delivered to houses. Since the objective is to minimize the cost of charging, and there are constraints on when the vehicles can charge, this problem can be formulated as an optimization problem. Minimizing a cost given constraints creates an optimization problem. An optimization solution approach called Ant Colony Optimization is used firstly to create energy-efficient routes and secondly to schedule the routes in a way that aims to minimize the cost of charging the electric-vehicle fleet. The electric-vehicle fleet in this work has different types of electric vehicles and different schedule constraints based on the tasks the vehicles complete. The ant colony approach makes good solutions in a relatively short period of time
A Provable Semi-Infinite Programming Approach for Solving Dynamic Nash Games
No full textMany engineering problems must account for the non-cooperative decisions and actions of multiple players. These problems can be modeled within a game-theoretic framework. The approach herein is to model problems as Nash games, convert them to semi-infinite programs, and leverage provable semi-infinite algorithms to solve the original problem. A particular algorithm that leverages off-the-shelf solvers is used to solve four low-dimensional benchmark problems successfully. Two types of linear quadratic dynamic games are then investigated: ones where each player’s problem is convex and ones where at least one player’s problem is nonconvex. Within each type, variations based on information structure, communication structure, number of players, and semi-infinite objective are considered. The selected algorithm in conjunction with MATLAB’s fmincon successfully solves all cases except the distributed communication case. The numerical solutions approximate theoretical solutions (when they are known) within approximately one percent. Run times varied problem to problem from seconds to days. Capture cases proved challenging as they correspond to singular solutions in optimal control
Effects of Elevation, Fire Age, Burn Severity, And Pre-Fire Vegetation on Post-Fire Sagebrush Songbird Communities
Get PDFThe Sagebrush Biome of western North America has been threatened by a combination of interconnected threats. In the Great and Columbian Basins, a large portion of sagebrush loss has been due to an increase in higher-severity wildfires, caused by invasive annual grasses. As sagebrush declines, numerous wildlife species that rely on the biome have experienced similar trends. Sagebrush-adapted songbirds such as sage thrasher, Brewer\u27s sparrow, and sagebrush sparrow as well as steppe-songbirds, have all experienced population declines. Currently, limited information exists as to how fire severity or recovery affect these songbirds or what resource managers can do to improve their outcomes post-fire.
To address these research and management needs, I conducted a large-scale observational study, where I examined how the abundance of sagebrush obligate and steppe-associated songbird species was influenced by fires that occurred in sagebrush habitats in northern Utah and southern Idaho. I conducted multi-species point counts within fire perimeters randomly stratified by elevation, fire age, and initial burn severity, as well as in unburned reference sites. Then, I modeled how each of these factors influenced songbird density. I found that sagebrush-adapted species were more abundant in reference areas, while steppe-associated species showed positive or neutral responses to fire. Within twenty-five years after a fire, many species occurred at similar densities to reference areas. Lower elevation sites and higher severity burns had more pronounced differences in avian densities compared to reference areas. Our results demonstrate that fire can result in a loss of available habitat for sagebrush songbirds, but that many of the species can return to pre-fire densities within twenty years, especially at higher elevations
Next, to assess whether pre-fire vegetation affects post-fire songbird abundance, I used remotely sensed data from the Rangeland Analysis Platform from the year before each fire occurred. I compared pre-fire vegetation cover to the densities of Brewer\u27s sparrows, green-tailed towhees, and the steppe-associated species from the previous chapter. I found that all species, except for horned lark, were more abundant in burned areas when pre-fire annual grass cover was lower. Sagebrush-associated species also occurred at higher densities when pre-fire shrub and perennial herbaceous cover were higher. Reducing annual grass cover before an area burns may lead to improved outcomes for many of these songbird species