New Evidential Reasoning-Rule Based Optimal Maintenance Time Determining Method Considering Dynamic Parameter Boundary

Identifying the optimal maintenance timing (OMT) of complex industrial equipment (CIE) is a highly effective way to maintain its stable operation and reduce unreasonable maintenance costs. Currently, determination of OMT for CIE involves the following two main issues: an accurate mathematical model is hard to establish and difficulty in guaranteeing the validity of OMT determination. Hence, in this article, a new OMT model considering cohesion factor has been developed based on the evidential reasoning rule (ER-rule), where dynamic boundaries are set for model parameters to determine the optimal solution more accurate and faster in the optimization process (CERr-DB). In CERr-DB, cohesion factor is introduced to represent the fusion of weight and reliability of evidence in the original ER-rule method to make it more interpretable. What is more, CERr-DB adopts a new output feedback mechanism to provide a dynamic boundary solution space for the optimization process of parameters, which effectively improves the search accuracy and rate. In an effort to quantitatively assess the effect of cohesion factors and dynamic boundaries on the properties of CIE, a sensitivity analysis based on the transparency and traceability of the CERr-DB methodology has been performed. A case study of the electric servo mechanism is carried out to validate the effectiveness of the new OMT model.