An application of systemic prediction evaluation parameters for neural network remaining useful life predictions models

PHM (Prognostic and Health Management) is a new concept, which is to ensure the normal operation of the complicated system, to achieve its functionality and reliability better. In this situation, the remaining useful life (RUL) prediction has aroused more and more concerns. Although the life prediction methods are many, there are not a set of systemic of evaluation parameters to evaluate the accuracy of the prediction method. The existing parameters have different limitations. Therefore, due to the shortage of the existing parameters, this paper presents the systemic prediction evaluation parameters to evaluate the prediction ability of the method. Then, some parameters are applied to the BP neural network in a power supply system of the evaluation.