11115 research outputs found
Sort by
Impact of Exposure Time on Optical-phase Measurements
No full textIn this paper, we explore the impact of exposure time on optical-phase measurements collected on light that has propagated through atmospheric-optical turbulence. We model the exposure time by phase averaging over a convective distance, and we quantify the associated impact of imposing an exposure time using the piston- and tilt-removed phase variance. We accomplish this analysis through the development of an analytic solution and wave-optics simulations. In turn, we show that the analytic solution and simulation results are in good agreement when Ucτ/D≲0.25, where Uc is the convective velocity, τ is the exposure time, and D is the aperture diameter. When Ucτ/D≳0.25, the analytic solution underestimates the piston- and tilt-removed phase variance relative to the simulation results, and we discuss these differences. This work, at large, informs wavefront sensing and adaptive-optics efforts, where either the wind speed is high, the system is on a high-speed platform, the beacon is on a high-speed platform, or the beacon signal is very faint thereby requiring long-exposure data collections
An Integrated Space Test Lexicon: A Taxonomy for the Integrated Test and Evaluation of Space Systems
Get PDFThe proposed Integrated Space Test Lexicon is intended to amalgamate the numerous definitions of integrated (IT or IT&E), development test (DT or DT&E), and operational test (OT or OT&E) into unified, service-wide definitions, aligned with the Space Test Enterprise Vision. Refining such definitions will help distill the core characteristics of these fundamental test types to first identify space system activities composing what is traditionally known as DT and OT, then to provide a means of how these activities fit into the IT paradigm and support space system development. In forging a common understanding of how DT and OT support space systems and capabilities, this lexicon will facilitate the foundation for an IT architecture, specifically the National Space Training and Testing Complex and the larger enterprise-level operational test and training infrastructure
Graph Theory Metrics for the Prioritization of Water Distribution Network Assets: A Case Study of Tyndall AFB, FL
No full textWater distribution networks, like other large infrastructure systems, must consider reliability and resilience efforts to resist and recover from failure with limited resources. Management of these assets requires a plan to prioritize the maintenance of system components, which are key to its reliability. Many current asset management practices for water distribution networks include only reactive strategies, such as fixing components after breaking, or on predetermined maintenance schedules, which might not correlate with a component’s condition or need. Asset management and graph theory principles are utilized in this study to evaluate a water distribution network and its vulnerabilities at Tyndall Air Force Base (AFB). We begin by estimating pipe condition indices and establishing the importance of each pipe through a set of graph-theoretic measures (e.g., pipe betweenness centrality and network efficiency). We next simulate two pipe failure scenarios: (1) a single pipe failure scenario, and (2) a cascading failure scenario. The outcomes of the second scenario reveal that a larger betweenness centrality equates to being a critical pipe in relation to overall system performance. Further, we estimate the risk associated with each pipe by combining pipe condition and network properties. We find that ∼7 km (or ∼6%) of pipes are at high risk, while the majority of the network’s pipes (∼79 km or ∼62%) are low risk. This study demonstrates how integrating condition indices with graph theory metrics could be employed to promote best practices in pipe maintenance and create a risk-informed framework for asset management
Random forests for detecting weak signals and extracting physical information: A case study of magnetic navigation
No full textIt has been recently demonstrated that two machine-learning architectures, reservoir computing and time-delayed feed-forward neural networks, can be exploited for detecting the Earth’s anomaly magnetic field immersed in overwhelming complex signals for magnetic navigation in a GPS-denied environment. The accuracy of the detected anomaly field corresponds to a positioning accuracy in the range of 10–40 m. To increase the accuracy and reduce the uncertainty of weak signal detection as well as to directly obtain the position information, we exploit the machine-learning model of random forests that combines the output of multiple decision trees to give optimal values of the physical quantities of interest. In particular, from time-series data gathered from the cockpit of a flying airplane during various maneuvering stages, where strong background complex signals are caused by other elements of the Earth’s magnetic field and the fields produced by the electronic systems in the cockpit, we demonstrate that the random-forest algorithm performs remarkably well in detecting the weak anomaly field and in filtering the position of the aircraft. With the aid of the conventional inertial navigation system, the positioning error can be reduced to less than 10 m. We also find that, contrary to the conventional wisdom, the classic Tolles–Lawson model for calibrating and removing the magnetic field generated by the body of the aircraft is not necessary and may even be detrimental for the success of the random-forest method
Responsible Machine Learning for United States Air Force Pilot Candidate Selection
No full textThe United States Air Force (USAF) continues to be plagued by a chronic pilot shortage, one that could be exacerbated by an accompanying shortfall in the commercial airlines. As a result, efforts have increased to alleviate this shortage by finding methods to reduce pilot training attrition. We contribute to these efforts by setting forth a decision support system (DSS) for pilot candidate selection using modern machine learning techniques. In view of the recent Responsible Artificial Intelligence Strategy published by the United States Department of Defense, this research leverages interpretable and explainable machine learning methods to create traceable and equitable models that may be responsibly and reliably governed. These models are used to regress candidates’ average merit assignment selection system scores based on information available for selection and prior to training. More specifically, using data provided by the USAF from 2010 to 2018, this paper develops and analyzes multiple interpretable models based on Gaussian Bayesian networks, as well as multiple black-box models rendered explainable by SHAP values and conformal prediction. A preferred pair of interpretable and explainable models is selected and embedded within a DSS for USAF pilot candidate selection boards: the Air Force Pilot Applicant Selection System. The utilization of this DSS is explored, the analyses it enables are discussed, and relevant USAF policymaking issues are examined
Residual Optical Absorption from Native Defects in CdSiP\u3csub\u3e2\u3c/sub\u3e Crystals
Get PDFCdSiP2 crystals are used in optical parametric oscillators to produce tunable output in the mid-infrared. As expected, the performance of the OPOs is adversely affected by residual optical absorption from native defects that are unintentionally present in the crystals. Electron paramagnetic resonance (EPR) identifies these native defects. Singly ionized silicon vacancies (V-Si) are responsible for broad optical absorption bands peaking near 800, 1033, and 1907 nm. A fourth absorption band, peaking near 630 nm, does not involve silicon vacancies. Exposure to 1064 nm light when the temperature of the CdSiP2 crystal is near 80K converts V-Si acceptors to their neutral and doubly ionized charge states (V0-Si and V2-Si , respectively) and greatly reduces the intensities of the three absorption bands. Subsequent warming to room temperature restores the singly ionized charge state of the silicon vacancies and brings back the absorption bands. Transitions responsible for the absorption bands are identified, and a mechanism that allows 1064 nm light to remove the singly ionized charge state of the silicon vacancies is proposed
An Assessment of Rules of Thumb for Software Phase Management, and the Relationship Between Phase Effort and Schedule Success
No full textIn the planning of a software development project, managers must estimate the amount of effort needed for distinct phases of activity. A number of rules of thumb exist in the literature to help the program manager in this task. However, very little work has been done to validate these rules of thumb. Applying least square models and Hotelling’s T 2 test, we evaluate these rules of thumb against a large database of Department of Defense projects. We determine that variability limits the simple application of any such rule. However, there are some worthy of closer attention, and we recommend adjustments for improved application. We also determine that projects which give extra attention to early phases experience less schedule growth. These findings were robust across developmental process type, military service, and project size
Standardization of Risk Classifications for Unmanned Space Vehicle Missions
Get PDFThis paper seeks to model risk classification levels (A-D) for 122 Space Vehicle programs. Models include multinomial logistic regression as well as random forest, a machine learning technique based on decision trees. We use independent variables (IVs) which are theoretically correlated to risk class for the regression and one random forest model. We then include all IVs and allow the random forest technique to use those which provide the most information on risk class before paring down the number of IVs to only 7. We show that the accuracy of predictions increases from 62% to 87% by using random forest and emphasize the adaptability of the models developed in code to new data
Toward space architecture resilience: a system-theoretic framework for analysis and design
No full textAs space-based systems have developed to provide numerous critical capabilities for defense applications, the systems designed to disrupt them have developed as well. Therefore, it is imperative to design, operate, and sustain space architectures for resilience. However, there are limited methodologies available which incorporate adversarial threats for the analysis and design of resilient space architectures. This paper recommends a System-Theoretic Process Analysis (STPA) framework to qualitatively analyze space architectures and identify design considerations, requirements, and constraints required for resilience. The study demonstrates the analysis framework on a notional remote sensing architecture subjected to a set of adversarial threats. The resulting framework provides (1) a process to assess space architectures for loss scenarios involving adversary threats and (2) a process to determine system behavioral constraints and mitigations to ensure system resilience
Federated Medical Scoring Systems
Get PDFFederated Learning (FL) is a recent framework of machine learning implementation that trains models on a distributed network of clients as opposed to housing and analyzing this data centrally. This has data communication and practical data privacy advantages, the latter of which is particularly attractive to the medical community where patient privacy is closely safeguarded. We apply FL to a family of sparse linear integer models called Medical Scoring Systems (MSSs). We create a novel methodology for creating these MSSs in a simulated federated environment that involves an lo constrained Logistic Regression (LR), loss-surface examination, and rounding procedure. We tested this methodology on two datasets from the University of California Irvine’s Machine Learning Repository, the Heart Disease and Mushroom datasets. We evaluated our federated MSSs to six other model architectures of varying interpretability, sparsity, and centrality. Ultimately, we succeeded in creating a methodology for a federated MSS that addresses medical privacy concerns and enhances interpretability whose performance was similar to the highest performing centralized model