A Systematic Approach for Developing Prognostic Algorithms on Large Complex Systems

Abstract

In order to apply prognostic and health management (PHM) techniques on a large scale, a systematic method and toolset are required. This paper describes the background, requirements, process and outputs of the PHM Pro tool, a Microsoft Access database designed to derive prognostic algorithms for the F-35 Joint Strike Fighter (JSF) Air System. The F-35 JSF is composed of over 2000 line replaceable components, each of which has the potential for functional failure and the application of automated prognostics. Using criteria agreed to by the customer, PHM Pro categorizes these 2000 items according to their need and applicability for prognostics. Once categorized, the tool walks the user through the steps to building the algorithm as it applies to the JSF architecture. These steps include signal collection, on-board preprocessing, failure progression trending, time to maintenance prediction, maturation and verification. The outputs of the tool are used by the JSF software development team to build the software code