Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Reduce order modeling”

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 865 records · Page 48
All sourcesNASA NTRSDOE OSTI
NASA NTRS2002

Probabilistic Study of Fluid Structure Interaction

Probabilistic CFD design is needed because we are asked to do more with less. To cost effectively accomplish the design task, we need to formally quantify the effect of uncertainties (variables) in the design. Probabilistic design is one effective method to formally quantify the effect of uncertainties. Our objective is to establish a revolutionary new early design process, by developing non-deterministic physics-based probabilistic design tools, which will include all the life cycle processes. Breakthroughs will be sought in speed, accuracy, intelligence, and usability of the system. This paper is concerned with the usefulness of parametric optimization method coupled with a Navier-Stokes analysis code for the aero-thermodynamic design of turbomachinery combustor liner. The interconnection between the CFD code and NESSUS codes facilitated the coupling between the thermal profiles and structural design. We have developed new concepts for reducing the computational cost of unsteady, three-dimensional, compressible aerodynamic analyses for multistage turbomachinery flows. The flow was modeled by the three-dimensional Favre-Reynolds-averaged Navier-Stokes equations using the k-epsilon turbulence closure, which was integrated using an implicit third-order upwind solver. The methodology developed in this paper is expected to lead to the design optimization of turbomachinery blades.

Gorla, Rama S. R.↗
NASA NTRS2003

Probabilistic Unsteady Aerodynamic Analysis

Probabilistic CFD design is needed because we are asked to do more with less. To cost effectively accomplish the design task, we need to formally quantify the effect of uncertainties (variables) in the design. Probabilistic design is one effective method to formally quantify the effect of uncertainties. Our objective is to establish a revolutionary new early design process, by developing non-deterministic physics-based probabilistic design tools, which will include all the life cycle processes. This work was concerned with the usefulness of parametric optimization method coupled with a Navier-Stokes analysis code for the aero-thermodynamic design of turbomachinery combustor liner. The interconnection between the CFD code and NESSUS codes will facilitate the coupling between the thermal profiles and structural design. We have developed new concepts for reducing the computational cost of unsteady, three-dimensional, compressible aerodynamic analyses for multistage turbomachinery flows. The flow was modeled by the three-dimensional Favre-Reynolds-averaged Navier-Stokes equations using the k-E turbulence closure, which was integrated using an implicit third-order upwind solver. The methodology developed in this work is expected to lead to the design optimization of turbomachinery blades.

Gorla, Rama S. R.↗
NASA NTRS2009

Measuring Massive Black Hole Binaries with LISA

The coalescence of two massive black holes produces gravitational waves (GWs) which can be detected by the space-based detector LISA. By measuring these waves, LISA can determine the various parameters which characterize the source. Measurements of the black hole masses and spins will provide information about the growth of black holes and their host galaxies over time. Measurements of a source's sky position and distance may help astronomers identify an electromagnetic counterpart to the GW event. The counterpart's redshift, combined with the GW-measured luminosity distance, can then be used to measure the Hubble constant and the dark energy parameter $w$. Because the potential science output is so high, it is useful to know in advance how well LISA can measure source parameters for a wide range of binaries. We calculate expected parameter estimation errors using the well-known Fisher matrix method. Our waveform model includes the physics of spin precession, as well as subleading harmonics. When these higher-order effects are not included, strong degeneracies between some parameters cause them to be poorly determined by a GW measurement. When precession and subleading harmonics are properly included, the degeneracies are broken, reducing parameter errors by one to several orders of magnitude.

Lang, Ryan N.↗
NASA NTRS2016

Soil Moisture Active Passive Mission L4_SM Data Product Assessment (Version 2 Validated Release)

During the post-launch SMAP calibration and validation (Cal/Val) phase there are two objectives for each science data product team: 1) calibrate, verify, and improve the performance of the science algorithm, and 2) validate the accuracy of the science data product as specified in the science requirements and according to the Cal/Val schedule. This report provides an assessment of the SMAP Level 4 Surface and Root Zone Soil Moisture Passive (L4_SM) product specifically for the product's public Version 2 validated release scheduled for 29 April 2016. The assessment of the Version 2 L4_SM data product includes comparisons of SMAP L4_SM soil moisture estimates with in situ soil moisture observations from core validation sites and sparse networks. The assessment further includes a global evaluation of the internal diagnostics from the ensemble-based data assimilation system that is used to generate the L4_SM product. This evaluation focuses on the statistics of the observation-minus-forecast (O-F) residuals and the analysis increments. Together, the core validation site comparisons and the statistics of the assimilation diagnostics are considered primary validation methodologies for the L4_SM product. Comparisons against in situ measurements from regional-scale sparse networks are considered a secondary validation methodology because such in situ measurements are subject to up-scaling errors from the point-scale to the grid cell scale of the data product. Based on the limited set of core validation sites, the wide geographic range of the sparse network sites, and the global assessment of the assimilation diagnostics, the assessment presented here meets the criteria established by the Committee on Earth Observing Satellites for Stage 2 validation and supports the validated release of the data. An analysis of the time average surface and root zone soil moisture shows that the global pattern of arid and humid regions are captured by the L4_SM estimates. Results from the core validation site comparisons indicate that "Version 2" of the L4_SM data product meets the self-imposed L4_SM accuracy requirement, which is formulated in terms of the ubRMSE: the RMSE (Root Mean Square Error) after removal of the long-term mean difference. The overall ubRMSE of the 3-hourly L4_SM surface soil moisture at the 9 km scale is 0.035 cubic meters per cubic meter requirement. The corresponding ubRMSE for L4_SM root zone soil moisture is 0.024 cubic meters per cubic meter requirement. Both of these metrics are comfortably below the 0.04 cubic meters per cubic meter requirement. The L4_SM estimates are an improvement over estimates from a model-only SMAP Nature Run version 4 (NRv4), which demonstrates the beneficial impact of the SMAP brightness temperature data. L4_SM surface soil moisture estimates are consistently more skillful than NRv4 estimates, although not by a statistically significant margin. The lack of statistical significance is not surprising given the limited data record available to date. Root zone soil moisture estimates from L4_SM and NRv4 have similar skill. Results from comparisons of the L4_SM product to in situ measurements from nearly 400 sparse network sites corroborate the core validation site results. The instantaneous soil moisture and soil temperature analysis increments are within a reasonable range and result in spatially smooth soil moisture analyses. The O-F residuals exhibit only small biases on the order of 1-3 degrees Kelvin between the (re-scaled) SMAP brightness temperature observations and the L4_SM model forecast, which indicates that the assimilation system is largely unbiased. The spatially averaged time series standard deviation of the O-F residuals is 5.9 degrees Kelvin, which reduces to 4.0 degrees Kelvin for the observation-minus-analysis (O-A) residuals, reflecting the impact of the SMAP observations on the L4_SM system. Averaged globally, the time series standard deviation of the normalized O-F residuals is close to unity, which would suggest that the magnitude of the modeled errors approximately reflects that of the actual errors. The assessment report also notes several limitations of the "Version 2" L4_SM data product and science algorithm calibration that will be addressed in future releases. Regionally, the time series standard deviation of the normalized O-F residuals deviates considerably from unity, which indicates that the L4_SM assimilation algorithm either over- or under-estimates the actual errors that are present in the system. Planned improvements include revised land model parameters, revised error parameters for the land model and the assimilated SMAP observations, and revised surface meteorological forcing data for the operational period and underlying climatological data. Moreover, a refined analysis of the impact of SMAP observations will be facilitated by the construction of additional variants of the model-only reference data. Nevertheless, the “Version 2” validated release of the L4_SM product is sufficiently mature and of adequate quality for distribution to and use by the larger science and application communities.

SMAP L4_SM↗
NASA NTRS2008

Calibrating the Helium Pressurization System for the Space Shuttle Liquid-Hydrogen Tank

Analysis of the results from the STS-114 tanking tests and subsequent launch called into question existing thermal and mass models of helium pressurization of the liquid hydrogen tank. This hydrogen tank, which makes up the bottom two-thirds of the External Tank, is pressurized prior to launch to avoid cavitation in the Shuttle Main Engine pumps. At about 2 minutes prior to launch, the main vent valve is closed, and pressurized helium flows into the tank ullage space to achieve set point pressure. As the helium gas cools, its pressure drops, calling for additional helium. Subsequent helium flows are provided in short, timed pulses. The number of pulses is taken as a rough leak indicator. An analysis of thermal models by Marshall Space Flight Center showed considerable uncertainty in the pressure-versus-time behavior of the helium ullage space and the ability to predict the number of pulses normally expected. Kennedy Space Center proposed to calibrate the dime-sized orifice, which together with valves, controls the helium flow quantity (Figure 1). Pressure and temperature sensors were installed to provide upstream and downstream measurements necessary to compute flow rate based on the orifice discharge coefficient. An assessment of flow testing with helium indicated an extremely costly use of this critical resource. In order to reduce costs, we proposed removing the orifices from each Mobile Launcher Platform (MLP) and asking Colorado Engineering Experiment Station Inc. (CEESI) to calibrate the flow. CEESI has a high-pressure air flow system with traceable flow meters capable of handling the large flow rates. However, literature research indicated that square-edged orifices of small diameters often exhibit significant hysteresis and nonrepeatability in the vicinity of choked or sonic flow. Fortunately, the MLP orifices behaved relatively well in testing (Figure 2). Using curve fitting of the air-flow data, in conjunction with ASME orifice modeling equations, a method of relating the helium mass flow to measured air flow data was obtained. This analysis showed that the highest uncertainty in flow occurred in the vicinity of the choking pressure ratio, as would be expected. In addition, analysis of typical flow pulses showed that most of the helium flow occurred either well below or well above this uncertain area. The final result is the ability to provide postlaunch estimates of helium mass flows that are within 1.5 percent of the actual value.

Source record↗
NASA NTRS1978

Filtering of non-linear instabilities

For Courant numbers larger than one and cell Reynolds numbers larger than two, oscillations and in some cases instabilities are typically found with implicit numerical solutions of the fluid dynamics equations. This behavior has sometimes been associated with the loss of diagonal dominance of the coefficient matrix. It is shown that these problems can be related to the choice of the spatial differences, with the resulting instability related to aliasing or nonlinear interaction. Appropriate filtering can reduce the intensity of these oscillations and possibly eliminate the instability. These filtering procedures are equivalent to a weighted average of conservation and nonconservation differencing. The entire spectrum of filtered equations retains a three point character as well as second order spatial accuracy. Burgers equation was considered as a model.

Khosla, P. K.↗
NASA NTRS1981

High resolution angular sensor

Specifications for the pointing stabilization system of the large space telescope were used in an investigation of the feasibility of reducing ring laser gyro output quantization to the sub-arc-second level by the use of phase locked loops and associated electronics. Systems analysis procedures are discussed and a multioscillator laser gyro model is presented along with data on the oscillator noise. It is shown that a second order closed loop can meet the measurement noise requirements when the loop gain and time constant of the loop filter are appropriately chosen. The preliminary electrical design is discussed from the standpoint of circuit tradeoff considerations. Analog, digital, and hybrid designs are given and their applicability to the high resolution sensor is examined. the electrical design choice of a system configuration is detailed. The design and operation of the various modules is considered and system block diagrams are included. Phase 1 and 2 test results using the multioscillator laser gyro are included.

Gneses, M. I.↗
NASA NTRS1986

Solvability condition for needle crystals at large undercooling in a nonlocal model of solidification

It is explicitly shown that, in a realistic model of diffusion-controlled dendritic solidification, Ivantsov's continuous family of steady-state needle crystals is destroyed by the addition of surface tension. The starting point is in the exact integro-differential equation for the one-sided model, in two dimensions, in a moving frame of reference. In the limit of large undercooling, where the range of the diffusion field is much smaller than the radius of curvature of the tip of the needle, this problem is reduced to a linear, inhomogeneous differential equation of infinite order. A solvability condition for this equation is derived and it is shown that solutions cease to exist for arbitrarily small but finite isotropic surface tension.

Caroli, B.↗
NASA NTRS1989

Control/structure interaction methods for space station power systems

The Structural Dynamics Research Corporation and the NASA Lewis Research Center have been working together to develop tools and methods for the analysis of control/structure interaction problems related to the space station power systems. Flexible modes of the solar arrays below 0.1 Hz, suggest that even for relatively slow control systems, the potential for control/structure interaction exists. The emphasis of the effort has been to develop tools which couple NASTRAN's powerful capabilities in structural dynamics with EASY5's powerful capabilities in control systems analysis. One product is an interface software package called CO-ST-IN for COntrol-STructure-INteraction. CO-ST-IN acts to translate data between NASTRAN and EASY5, facilitating the analysis of complex coupled problems. Interfaces to SDRC I-DEAS and MATRIXx are also offered. Beside transferring standard modal information, CO-ST-IN implements a number of advanced methods. These include a modal ordering algorithm that helps eliminate uncontrollable or unobservable modes from the analysis, an implementation of the more accurate mode acceleration algorithm for recovery of element forces and stresses directly in EASY5 and an implementation of fixed interface modes in NASTRAN, which reduces the error in the closed-loop model due to the use of truncated mode sets.

Blelloch, Paul↗
NASA NTRS1993

Modeling of topographic effects on Antarctic sea ice using multivariate adaptive regression splines

The role of seafloor topography in the spatial variations of the southern ocean sea ice cover as observed (every other day) by the Nimbus 7 scanning multichannel microwave radiometer satellite in the years 1980, 1983, and 1984 is studied. Bottom bathymetry can affect sea ice surface characteristics because of the basically barotropic circulation of the ocean south of the Antarctic Circumpolar current. The main statistical tool used to quantify this effect is a local nonparametric regression model of sea ice concentration as a function of the depth and its first two derivatives in both meridional and zonal directions. First, we model the relationship of bathymetry to sea ice concentration in two sudy areas, one over the Maud Rise and the other over the Ross Sea shelf region. The multiple correlation coefficient is found to average 44% in the Maud Rise study area and 62% in the Ross Sea study area over the years 1980, 1983, and 1984. Second, a strategy of dividing the entire Antarctic region into an overlapping mosaic of small areas, or windows is considered. Keeping the windows small reduces the correlation of bathymetry with other factors such as wind, sea temperature, and distance to the continent. We find that although the form of the model varies from window to window due to the changing role of other relevant environmental variables, we are left with a spatially consistent ordering of the relative importance of the topographic predictors. For a set of three representative days in the Austral winter of 1980, the analysis shows that an average of 54% of the spatial variation in sea ice concentration over the entire ice cover can be attributed to topographic variables. The results thus support the hypothesis that there is a sea ice to bottom bathymetry link. However this should not undermine the considerable influence of wind, current, and temperature which affect the ice distribution directly and are partly responsible for the observed bathymetric effects.

De Veaux, Richard D.↗
NASA NTRS2002

Error-Based Design Space Windowing

Windowing of design space is considered in order to reduce the bias errors due to low-order polynomial response surfaces (RS). Standard design space windowing (DSW) uses a region of interest by setting a requirement on response level and checks it by a global RS predictions over the design space. This approach, however, is vulnerable since RS modeling errors may lead to the wrong region to zoom on. The approach is modified by introducing an eigenvalue error measure based on point-to-point mean squared error criterion. Two examples are presented to demonstrate the benefit of the error-based DSW.

Papila, Melih↗
NASA NTRS2002

Implementation of the NCAR Community Land Model (CLM) in the NASA/NCAR finite-volume Global Climate Model (fvGCM)

In this study, the NCAR CLM version 2.0 land-surface model was integrated into the NASA/NCAR fvGCM. The CLM was developed collaboratively by an open interagency/university group of scientists and based on well-proven physical parameterizations and numerical schemes that combine the best features of BATS, NCAR-LSM, and IAP94. The CLM design is a one-dimensional point model with 1 vegetation layer, along with sub-grid scale tiles. The features of the CLM include 10-uneven soil layers with water, ice, and temperature states in each soil layer, and five snow layers, with water flow, refreezing, compaction, and aging allowed. In addition, the CLM utilizes two-stream canopy radiative transfer, the Bonan lake model and topographic enhanced streamflow based on TOPMODEL. The DAO fvGCM uses a genuinely conservative Flux-Form Semi-Lagrangian transport algorithm along with terrain- following Lagrangian control-volume vertical coordinates. The physical parameterizations are based on the NCAR Community Atmosphere Model (CAM-2). For our purposes, the fvGCM was run at 2 deg x 2.5 deg horizontal resolution with 55 vertical levels. The 10-year climate from the fvGCM with CLM2 was intercompared with the climate from fvGCM with LSM, ECMWF and NCEP. We concluded that the incorporation of CLM2 did not significantly impact the fvGCM climate from that of LSM. The most striking difference was the warm bias in the CLM2 surface skin temperature over desert regions. We determined that the warm bias can be partially attributed to the value of the drag coefficient for the soil under the canopy, which was too small resulting in a decoupling between the ground surface and the canopy. We also discovered that the canopy interception was high compared to observations in the Amazon region. A number of experiments were then performed focused on implementing model improvements. In order to correct the warm bias, the drag coefficient for the soil under the canopy was considered a function of LAI (Leaf Area Index). Analysis of the results revealed that there was a substantial impact, and the warm and dry bias in the CLM2 was significantly reduced. For the interception scheme, the canopy throughfall was increased to allow for more infiltration of precipitation into the soil, resulting in increased low-level moisture and a decrease in the interception loss ratio (canopy evaporation to precipitation).

Radakovich, Jon D.↗
NASA NTRS2000

Higher Order Modeling in Hybrid Approaches to the Computation of Electromagnetic Fields

Higher order geometry representations and interpolatory basis functions for computational electromagnetics are reviewed. Two types of vector-valued basis functions are described: curl-conforming bases, used primarily in finite element solutions, and divergence-conforming bases used primarily in integral equation formulations. Both sets satisfy Nedelec constraints, which optimally reduce the number of degrees of freedom required for a given order. Results are presented illustrating the improved accuracy and convergence properties of higher order representations for hybrid integral equation and finite element methods.

Wilton, Donald R.↗
NASA NTRS2015

Model-Based Systems Engineering in Concurrent Engineering Centers

Concurrent Engineering Centers (CECs) are specialized facilities with a goal of generating and maturing engineering designs by enabling rapid design iterations. This is accomplished by co-locating a team of experts (either physically or virtually) in a room with a focused design goal and a limited timeline of a week or less. The systems engineer uses a model of the system to capture the relevant interfaces and manage the overall architecture. A single model that integrates other design information and modeling allows the entire team to visualize the concurrent activity and identify conflicts more efficiently, potentially resulting in a systems model that will continue to be used throughout the project lifecycle. Performing systems engineering using such a system model is the definition of model-based systems engineering (MBSE); therefore, CECs evolving their approach to incorporate advances in MBSE are more successful in reducing time and cost needed to meet study goals. This paper surveys space mission CECs that are in the middle of this evolution, and the authors share their experiences in order to promote discussion within the community.

systems engineering↗
NASA NTRS2015

Model-Based Systems Engineering in Concurrent Engineering Centers

Concurrent Engineering Centers (CECs) are specialized facilities with a goal of generating and maturing engineering designs by enabling rapid design iterations. This is accomplished by co-locating a team of experts (either physically or virtually) in a room with a narrow design goal and a limited timeline of a week or less. The systems engineer uses a model of the system to capture the relevant interfaces and manage the overall architecture. A single model that integrates other design information and modeling allows the entire team to visualize the concurrent activity and identify conflicts more efficiently, potentially resulting in a systems model that will continue to be used throughout the project lifecycle. Performing systems engineering using such a system model is the definition of model-based systems engineering (MBSE); therefore, CECs evolving their approach to incorporate advances in MBSE are more successful in reducing time and cost needed to meet study goals. This paper surveys space mission CECs that are in the middle of this evolution, and the authors share their experiences in order to promote discussion within the community.

Iwata, Curtis↗
NASA NTRS2016

Recovery of Large Angular Scale CMB Polarization for Instruments Employing Variable-Delay Polarization Modulators

Variable-delay Polarization Modulators (VPMs) are currently being implemented in experiments designed to measure the polarization of the cosmic microwave background on large angular scales because of their capability for providing rapid, front-end polarization modulation and control over systematic errors. Despite the advantages provided by the VPM, it is important to identify and mitigate any time-varying effects that leak into the synchronously modulated component of the signal. In this paper, the effect of emission from a 300 K VPM on the system performance is considered and addressed. Though instrument design can greatly reduce the influence of modulated VPM emission, some residual modulated signal is expected. VPM emission is treated in the presence of rotational misalignments and temperature variation. Simulations of time-ordered data are used to evaluate the effect of these residual errors on the power spectrum. The analysis and modeling in this paper guides experimentalists on the critical aspects of observations using VPMs as front-end modulators. By implementing the characterizations and controls as described, front-end VPM modulation can be very powerful for mitigating 1/ f noise in large angular scale polarimetric surveys. None of the systematic errors studied fundamentally limit the detection and characterization of B-modes on large scales for a tensor-to-scalar ratio of r= 0.01. Indeed, r less than 0.01 is achievable with commensurately improved characterizations and controls.

Miller, N. J.↗
NASA NTRS2017

Computational Assessment of a 3-Stage Axial Compressor Which Provides Airflow to the NASA 11- by 11-Foot Transonic Wind Tunnel, Including Design Changes for Increased Performance

A 24 foot diameter 3-stage axial compressor powered by variable-speed induction motors provides the airflow in the closed-return 11- by 11-Foot Transonic Wind Tunnel (11-Foot TWT) Facility at NASA Ames Research Center at Moffett Field, California. The facility is part of the Unitary Plan Wind Tunnel, which was completed in 1955. Since then, upgrades made to the 11-Foot TWT such as flow conditioning devices and instrumentation have increased blockage and pressure loss in the tunnel, somewhat reducing the peak Mach number capability of the test section. Due to erosion effects on the existing aluminum alloy rotor blades, fabrication of new steel rotor blades is planned. This presents an opportunity to increase the Mach number capability of the tunnel by redesigning the compressor for increased pressure ratio. Challenging design constraints exist for any proposed design, demanding the use of the existing driveline, rotor disks, stator vanes, and hub and casing flow paths, so as to minimize cost and installation time. The current effort was undertaken to characterize the performance of the existing compressor design using available design tools and computational fluid dynamics (CFD) codes and subsequently recommend a new compressor design to achieve higher pressure ratio, which directly correlates with increased test section Mach number. The constant cross-sectional area of the compressor leads to highly diffusion factors, which presents a challenge in simulating the existing design. The CFD code APNASA was used to simulate the aerodynamic performance of the existing compressor. The simulations were compared to performance predictions from the HT0300 turbomachinery design and analysis code, and to compressor performance data taken during a 1997 facility test. It was found that the CFD simulations were sensitive to endwall leakages associated with stator buttons, and to a lesser degree, under-stator-platform flow recirculation at the hub. When stator button leakages were modeled, pumping capability increased by over 20 of pressure rise at design point due to a large reduction in aerodynamic blockage at the hub. Incorporating the stator button leakages was crucial to matching test data. Under-stator-platform flow recirculation was thought to be large due to a lack of seals. The effect of this recirculation was assessed with APNASA simulations recirculating 0.5, 1, and 2 of inlet flow about stators 1 and 2, modeled as axisymmetric mass flux boundary conditions on the hub before and after the vanes. The injection of flow ahead of the stators tended to re-energize the boundary layer and reduce hub separations, resulting in about 3 increased stall margin per 1 of inlet flow recirculated. In order to assess the value of the flow recirculation, a mixing plane simulation of the compressor which gridded the under-stator cavities was generated using the ADPAC CFD code. This simulation indicated that about 0.65 of the inlet flow is recirculated around each shrouded stator. This collective information was applied during the redesign of the compressor. A potential design was identified using HT0300 which improved overall pressure ratio by removing pre-swirl into rotor 1, replacing existing NASA 65 series rotors with double circular arc sections, and re-staggering rotors and the existing stators. The performance of the new design predicted by APNASA and HT0300 is compared to the existing design.

Turbomachinery↗
NASA NTRSDate not supplied

Early Information Parameter-Set Analysis for Satellite Close Approaches using Machine Learning

Understanding orbital mechanics is essential in space flight and navigation applications, and leveraging modern force models for flight path projection remains an important aspect in space mission design and operation. However, force models do not capture all the dynamics or perturbations in the space environment and thus are subject to errors in predicting the state vectors. The further out the predicted miss distance between spacecraft is from the time of closest approach (TCA), the larger the propagated errors in the predicted miss distance at TCA is. The dependency on these force models for spacecraft flight state prediction calls for a more reliable method that can quantify, or even reduce, these propagated errors. With recent advances in the field artificial intelligence, specifically in machine and deep learning algorithms, a model that implements these approaches can improve on the modern force model approach. The goal for this work is to provide an early-information decision-making threshold, in order to prioritize risk assessment implementation, given the ongoing increase of space objects. In analyzing the relationship of several parameters from conjunction data messages(CDMs) and solar information, early information becomes viable in miss distance prediction with unsupervised learning techniques, which learn the parameters that are linked together with miss distance and probability of collision (Pc) variables. Another approach implemented for identifying relationships within CDMs is supervised learning, in which a shallow neural network binary classifier learns to distinguish events with Pc values¡108. These parameters detected in the unsupervised process are then applied to a regression neural network, which predicts the miss distance at TCA for a specific event within a given uncertainty bound. For the regression neural network, a Long Short Term Memory (LSTM) neural network is implemented, which yields memory about each time step in an event. Using an LSTM network, the model learns to predict miss distance within 0.2km of the value measured at TCA. Although there is a limited amount of "close miss" data to train a network, the network learns to associate parameters, like large energy dissipation rates with the secondary object, with an elevated Pc

Brianna I. Robertson↗