Testability Design and Optimization of Satellite Based on Unified Multidisciplinary Modeling

Satellite systems，as complex engineering entities that span extensive domains and demand stringent precision，are currently constrained by traditional physical testing methods in testability design and verification，where critical system characteristics often remain unverifiable and fault detection/isolation processes incur prohibitively high costs. To overcome these limitations，a systematic methodology for digital testability design and optimization based on unified multidisciplinary modeling is developed. First，multi-signal flow graph theory is investigated to analyze the testability modeling capabilities of unified modeling languages，enabling testability parameters and test nodes to be systematically integrated for unambiguous mapping between system functional models and testability models. Second， component interconnections within the testability model are analyzed to derive fault-test correlations，from which correlation matrices are generated，while static analysis techniques are employed to identify system vulnerabilities. Third， optimal test point selection is implemented for efficient fault detection and isolation，and the AO^* heuristic algorithm is applied to optimize test sequences，thereby minimizing testing costs. Experimental validation using representative satellite power supply equipment demonstrates that this model-driven framework significantly enhances testing efficiency and accuracy while reducing implementation costs，providing a systematic solution for testability engineering in complex satellite systems.