Reliability modeling for failure correlations to a CNC lathe subsystem based on Pagerank Algorithm
Dongwei Gu,Jin Guo,Wenbo Han,Song Gao,Xilu Zhao,Zhen Xu
DOI: https://doi.org/10.1080/08982112.2023.2208664
2023-05-09
Quality Engineering
Abstract:The modeling of Computer Numerical Control (CNC) lathes has been constrained by various issues, such as the use of single methods, inaccurate model parameters, and inadequate consideration of subsystem failures. To overcome these challenges, this study proposes a reliability modeling method based on the Pagerank Algorithm for analyzing failure correlations in CNC lathe subsystems. The proposed method involves analyzing the failure correlation between subsystems through failure transfer-directed graph and representing the relationship between subsystems using an adjacency matrix. Based on the fault correlation between subsystem, a reliability model of the CNC lathe subsystem is established, and the Pagerank Algorithm is utilized to optimize the influence degree of each subsystem. To validate the proposed method, field use failure information of CNC lathe is utilized as an example, and the above theoretical modeling is verified and analyzed. Finally, the reliability model based on failure correlation is verified using the D-test value, relative error method, and accurate factor. In comparison to the reliability model without considering failure correlation, the proposed model results in a small relative error and more accurate value.
engineering, industrial,statistics & probability