A Framework for System Design Optimization Based on Maintenance Scheduling With Prognostics and Health Management

Abstract

The optimal maintenance scheduling of systems with degrading components is highly coupled with the design of the system and various uncertainties associated with the system, including the operating conditions, the interaction of different degradation profiles of various system components, and the ability to measure and predict degradation using prognostics and health management (PHM) technologies. Due to this complexity, designers need to understand the correlations and feedback between the design variables and lifecycle parameters to make optimal decisions. A framework is proposed for the high level integration of design, component degradation, and maintenance decisions. The framework includes constructing screening models for rapid design evaluation, defining a multi-objective robust optimization problem, and using sensitivity studies to compare trade-offs between different design and maintenance strategies. A case example of power plant condenser is used to illustrate the proposed framework and advise how designers can make informed comparisons between different design concepts and maintenance strategies under highly uncertain lifecycle conditions.