Filtering and prediction techniques for model-based prognosis and uncertainty management

Managing and reducing prognostic uncertainty is of significant importance to the success of PHM applications. The focus of prognosis uncertainty management is to identify and manage the reducible uncertainties by applying available data using appropriate uncertainty management algorithms. Particularly for dynamic model-based systems, opportunities exist to apply nonlinear filtering to provide a systematic way of dealing with the propagation of system damage at some future time, whenever imprecise diagnostic information is obtained. The goal of this paper is to present a foundation for prediction and filtering of the failure process using nonlinear prognostic models and filters, and illustrate how prognostic uncertainties are addressed within three types of filtering frameworks, namely the exact filtering, particle filtering and multiple-model filtering. Examples and illustrative simulation results are provided.