A Parameter Ontology for Simulation Models of Deep Space Habitats
Year: 2024
Type: article
Abstract: Deep space habitat operations will depend less on ground support than in low Earth orbit because of increasing communication delays. For the crew to be more autonomous, some of the resources dedicated to anomaly response mechanisms must be transferred from the ground to the habitat. Simulation models replicating the behavior of the system are examples of such resources as they can inform decisions to select mitigation strategies. Multiple models are typically developed for a single spacecraft as many domains of expertise and stakeholders are involved. Enabling simulation onboard a habitat will therefore require integration efforts to provide system-wide capabilities. Specifically, model parameters are model components that must be regularly calibrated to accurately represent the system state during operation. Depending on the type of simulation, these can differ greatly in their nature, thereby making difficult the tasks of creating, reading, updating, and deleting parameters. There is no standard specifying how to represent simulation parameters in a multimodel setting. Model ontologies have been shown to support integration, but their representation of model parameters is usually limited, i.e., the required type of parameter attributes and their hierarchical relationships are unknown. For example, quantifying the uncertainties in parameter values is essential in deep space operations but it is unclear how such information should be formally represented. In this article, we propose a parameter ontology extending existing standards with attributes enabling probabilistic parameter identification. This article is validated by comparing related work with the responses of the proposed ontology to competency questions and by evaluating its consistency, conciseness, completeness, and expandability.
Authors: Nicolas Gratius, Mario Bergés, Burcu Akinci
Institution: Carnegie Mellon University
Language: English
Subfield: Control and Systems Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: hybrid
Estimated APC paid: $2,495
OpenAlex ID: W4401452424