Research on Key Component Identification Methods for Multi-component Sorting System Under Cognitive Uncertainty

Sorting is a crucial part of logistics distribution, and identifying the key components that will have a significant impact on the multi-component sorting system once damaged is a necessary means to ensure the effective operation of the multi-component sorting system and maintain the stability of the entire logistics system. In the field of reliability, key components are identified by calculating component importance. However, the traditional importance is based on the two-state system, and it assumes that the degradation data of the system and components are completely accurate. These assumptions are inconsistent with the reality that the degradation of the multi-component sorting system is continuous, resulting in multiple working states. Furthermore, the degradation parameters are fuzzy due to cognitive uncertainty throughout this process. In response to this situation, this paper uses evidence theory to quantify the cognitive uncertainty and employs a Markov model to construct a component degradation model of the multi-state system. Building upon this foundation, the Birnbaum importance, Integrated Importance Measure (IIM) importance, Fussell-Vesely (FV) importance, and Risk Achievement Worth (RAW) importance are extended and applied to identify key components in the multi-component sorting system. This provides recommendations for selecting importance in different scenarios.