Abstract: In order to supplement the influence of component-related failures on the reliability allocation results of a system, a new reliability allocation method is proposed. The factors affecting system reliability allocation are divided into complexity, technical levels, environmental conditions, maintainability, operating time, and cost sensitivity, which are used to determine the weights of component reliability allocation. Among them, the first three factors are regarded as objective factors, quantitatively described and determined their values; while the last three factors are regarded as subjective factors, and their comprehensive scores are characterized by using an LNNWMM operator. A system reliability allocation model is established by the Copula function. A THP6513 heavy-duty CNC horizontal milling and boring machine spindle system is taken as an example to compare the reliability distribution results of subsystems with component-related failures and independent failures. Results show that the reliability of the six subsystems considering the failure correlation is 0.9720, 0.9835, 0.9829, 0.9778, 0.9819 and 0.9798, respectively, which are slightly lower than the reliability of each subsystem with independent component failures. It can reduce the manufacturing and maintenance costs of products.

Key words: reliability allocation, complexity, LNNWMM (linguistic neutrosophic numbers weight Muirhead mean operator), Copula function