Hierarchical Model-Based Approach to Testability Modeling and Analysis for PHM of Aerospace Systems

Abstract

A prognostics and health management (PHM) technique has been developed and applied to a variety of safety-critical aerospace systems. The PHM performance relies highly on test data, which conveys relevant system health information, and design for testability (DFT) developed concurrently with system design is thus of great importance to PHM performance. The testability model is the basis for testability analysis and design. To address the problems that traditional testability models did not include such as any quantitative testability information that could not describe fault-evolution test dependency, a novel model referred to as a quantified uncertainty hierarchical model is presented. In the model, fault-test dependency was described through quantified directed graph and fault attributes; test attributes and propagation attributes were assigned to nodes and directed edges in the form of probability, fuzziness, and uncertainty at the system level. And at component level, the physics of the failure model or extended failure modes, mechanisms, and effects analysis (FMMEA) were used to construct fault-evolution test dependency. The fault/fault-evolution test dependency was represented by a binary dependency matrix, based on the testability analysis for PHM, which can be realized when it is combined with quantitative testability information. Two cases were presented to demonstrate the proposed model for a missile control system and an aeroengine. Application analysis shows the proposed model is feasible and effective, and this approach can be used for testability modeling and PHM analysis of any system.