Metrics Library for Prognostics Performance Evaluation(ARC-17898-1)

The Metrics Library for Prognostics Performance Evaluation is a suite of algorithms implemented in MATLAB for evaluating the performance of prognostics algorithms that predict the remaining useful life (RUL) of engineering systems. It includes alpha-lambda, beta metrics that compare the ground truth RUL against continuous predictions; these predictions can either be point-value or be a probability distribution (that depicts the uncertainty regarding the prediction of RUL). The library can be used for a variety of applications and performs different types of evaluations depending upon how the ground truth and the predictions are available. The library is also suitable for a variety of prognostics algorithms.