ADEPS: a methodology for designing prognostic applications

Aizpurua, Jose Ignacio and Catterson, Victoria M.; Eballard, Ioana and Bregon, Anibal, eds. (2016) ADEPS: a methodology for designing prognostic applications. In: Proceedings of the Third European Conference of the Prognostics and Health Management Society 2016. PHM Society, ESP, pp. 86-100. ISBN 9781936263219

[thumbnail of Aizpurua-Catterson-PHME2016-ADEPS-a-methodology-for-designing-prognostic-applications]
Preview
Text. Filename: Aizpurua_Catterson_PHME2016_ADEPS_a_methodology_for_designing_prognostic_applications.pdf
Final Published Version
License: Creative Commons Attribution 4.0 logo

Download (553kB)| Preview

Abstract

Prognostics applications predict the future evolution of an asset under study, by diagnosing the actual health state and modeling the future degradation. Due to rapidly growing interest in prognostics, different prediction techniques have been developed independently without a consistent and systematic design. In this paper we formalize the prognostics design process with a novel methodology entitled ADEPS (Assisted Design for Engineering Prognostic Systems). ADEPS combines prognostics concepts with model-based safety assessment, criticality analysis, knowledge engineering and formal verification approaches. The main activities of ADEPS include synthesis of the safety assessment model from the design model, prioritization of the system failure modes, systematic prognostics model selection and verification of the adequacy of the prognostics results with respect to design requirements. By linking system-level safety assessment models and prognostics results, design and safety models are updated with online information about different failure modes. This step enables system-level health assessment including prognostics predictions of different failure modes. The end-to-end application of the methodology for the design and evaluation of a power transformer demonstrates the benefits of the proposed approach including reduced design time and effort, complete consideration of prognostics algorithms and updated system-level health assessment.