A maintenance support framework based on dynamic reliability and remaining useful life

Abstract Traditional maintenance support methods provide the maintenance strategy of equipment based on failure data from a number of similar equipment, without taking into account the specificity of degradation process. In this paper, a maintenance support framework based on dynamic reliability (DR) assessment and remaining useful life (RUL) prediction is developed for enhancing the practical maintenance activities from both statistical and individual perspective. First, a multi-factor fused DR assessment method is defined based on state offset degree, which can provide a more accurate DR result from the individual perspective. Second, a modified similarity-based RUL prediction method is proposed to improve prediction accuracy and reduce late prediction from the statistical perspective. In the end, a maintenance support model is defined based on RUL prediction and DR assessment, which can help managers to generate an effective maintenance strategy. A case study is provided to verify the effectiveness of the proposed method.