A Framework for Prognostics Reasoning

Open Access

Abstract

The use of system data to make predictions about the future system state commonly known as prognostics is a rapidly developing field. Prognostics seeks to build on current diagnostic equipment capabilities for its predictive capability. Many military systems including the Joint Strike Fighter (JSF) are planning to include on-board prognostics systems to enhance system supportability and affordability. Current research efforts supporting these developments tend to focus on developing a prognostic tool for one specific system component. This dissertation research presents a comprehensive literature review of these developing research efforts. It also develops presents a mathematical model for the optimum allocation of prognostics sensors and their associated classifiers on a given system and all of its components. The model assumptions about system criticality are consistent with current industrial philosophies. This research also develops methodologies for combine sensor classifiers to allow for the selection of the best sensor ensemble.