Title:
A Graph-Based Methodology for Model Inconsistency Identification and Robust Architecture Exploration and Analysis
A Graph-Based Methodology for Model Inconsistency Identification and Robust Architecture Exploration and Analysis
Author(s)
Duca, Ruxandra
Advisor(s)
Mavris, Dimitri N.
Editor(s)
Collections
Supplementary to
Permanent Link
Abstract
The rise in complexity in aircraft design and the move towards non-conventional architectures lead to errors discovered late when changes are costly. A leading cause is the distributed design with isolated but interdependent models, which makes it difficult to maintain a consistent set of assumptions. Several gaps were identified, then a methodology was proposed to (1) to define a novel, internally feasible candidate architecture, (2) ensure that external analysis models are consistent with it, and (3) systematically extract cross-tool dependencies for multi-disciplinary analysis setup.
In the first step, Model-Based Systems Engineering was leveraged to create a formal descriptive model of a baseline architecture. For this, an interface-based ontology was formulated using rules about component terminals and a standardized set of interactions. Incremental exploration was then enabled by developing a query-and-action process to find elements that must be added or removed after a local component replacement. The process was demonstrated by sequentially electrifying subsystems of a conventional baseline, resulting in numerous changes and restoring the system’s internal feasibility.
In the second step, the application of inconsistency detection methods was enabled by automating the search for semantic overlap between analysis models and the central descriptive model. For this, data from the two was encoded into labeled digraphs and an algorithm was used to find the maximum common subgraph. It was demonstrated between the electrified candidate architecture from the first step and a conventional aircraft model as seen by an analysis tool. After finding the equivalent elements, the inconsistency detection method was demonstrated.
The last step leveraged the results of the first two: analysis tools linked to a cross-disciplinary descriptive view of the whole system. Using the central model as an intermediary, cross-tool constraints were extracted, even when the relevant parameters were not exposed as inputs or outputs. This was demonstrated between the analysis model in the second step and a localized thermal model.
With a formal, cross-disciplinary view of the candidate architecture and a set of properly configured tools and cross-tool constraints, this methodology enables the exploration of subsystem architectures during preliminary design with less effort than current methods, and with the prospective of fewer errors being discovered in later stages of design.
Sponsor
Date Issued
2022-05-03
Extent
Resource Type
Text
Resource Subtype
Dissertation