NASA is responding to the growing interest in, and capabilities of, small satellites for science applications with an increasing number and frequency of Announcements of Opportunity for small satellite space missions. Estimating the probability that these mission concepts will fit within the small cost caps of these opportunities is largely driven by the probability that one of the burgeoning number of small satellite providers will be able to meet the payload's accommodation requirements within the budget for the spacecraft. JPL has collected a database containing technical specifications and cost of commercially available Smallsat buses across various vendors. The primary purpose of the database is for use in JPL's Team X architecture studies to inform cost estimates of a spacecraft bus which fits the customer's technical requirements for their payload and mission. Customer needs are often unique and don't align perfectly with an off-the-shelf commercial spacecraft bus, which motivates the need to develop a cost model across the continuous technical parameter space. Al's Bus Cost Distribution Estimator (ABCDE) uses Gaussian process regression (GPR) to predict commercial Smallsat spacecraft bus cost based on a subset of a customer's technical requirements (payload mass, payload power, delta V, pointing control, and downlink rate). GPR is implemented in ABCDE as a Bayesian method which fits an implied multivariate regression on the technical parameters and uses kriging to intentionally “overfit” the residuals. Overfitting the residuals allows cost estimates to collapse in uncertainty closer to the data points while maintaining larger uncertainty intervals in regions of parameter space with fewer data records. The data used to fit this model is sensitive and represents cost estimates for off-the-shelf commercial buses. GPR simultaneously protects the sensitivity of the database and uses the sparse nature of the database to account for uncertainty in cost in a useful way. For a given set of customer technical requirements, the tool provides a cost estimate distribution, the percentiles of which can be interpreted as a confidence level of finding a commercial bus under a specified cost cap. ABCDE dramatically pushes the boundaries of spacecraft cost estimation models due to its Bayesian methodology (accounting for the maximum uncertainty in the underlying regression), the mathematically advanced kriging methodology, and the novelty of its application in Team X architecture trade studies.


    Zugriff

    Zugriff prüfen

    Verfügbarkeit in meiner Bibliothek prüfen

    Bestellung bei Subito €


    Exportieren, teilen und zitieren



    Titel :

    Gaussian Process Regression Method for Costing SmallSat Bus Capabilities


    Beteiligte:
    Hooke, Melissa (Autor:in) / Fleischer, Sam (Autor:in) / Nash, Alfred (Autor:in) / Austin, Alex (Autor:in)


    Erscheinungsdatum :

    05.03.2022


    Format / Umfang :

    1909785 byte




    Medientyp :

    Aufsatz (Konferenz)


    Format :

    Elektronische Ressource


    Sprache :

    Englisch



    Gaussian Process Regression Method for Costing SmallSat Bus Capabilities

    Austin, Alex / Nash, Alfred / Fleischer, Sam et al. | NTRS | 2022




    SmallSat reliability

    Sheldon, D.J. | NTRS | 2017


    SmallSat Database

    Petropulos, Dolores / Bittner, David / Murawski, Robert et al. | NTRS | 2015