Formulation of Prognostics Requirements

Abstract

This paper proposes a methodology for formulating prognostics requirements for designers of electronic prognostics (EP)-enabled systems. A usefulness criterion is introduced that enables the specification of admissible uncertainty bounds on measurements of systems' health/degradation parameters. Keeping the uncertainty of health estimation below the admissible levels assures more accurate individual remaining useful life (RUL) estimations than those based on traditional population average time-to-failure. The proposed methodology is demonstrated using the well-known cumulative damage model that was extended to incorporate individual degradation data.