AHP‐IFM Target: An Innovative Method to Define Reliability Target in an Aerospace Prototype Based on Analytic Hierarchy Process

Abstract

Reliability target definition is a crucial aspect of any reliability analysis. In literature, there are two types of analysis. The first one, called ‘bottom‐up’, goes back to the system's target using data of units through a fault tree analysis. Reliability data of components could be only partially available, particularly in the case of innovative systems. In the second type of analysis, called ‘top‐down’, starting from similar systems, the target of each unit is defined, by applying allocation techniques. Also, in this case, reliability data of similar systems might not be available, and the choice of the most appropriate technique could be tricky. The purpose of the present research is to combine the advantages of both usual approaches. The newly developed approach is based on the integrated factors method, whose values are adjusted trough a multicriteria method, the analytic hierarchy process, depending on the importance of each factor and each unit. The innovation of the proposed model consists in its dynamism, as most of the literature methods use constant weights for the factors involved in reliability allocation. No method takes into account the assignment of a different level of significance (weight) to different units of the system, simultaneously with the considered factors. The developed approach has been applied on an aerospace prototype system. The results show the goodness of the new method and its ability to overcome the problems noted in literature. Copyright © 2017 John Wiley & Sons, Ltd.

A reliability allocation method of mechanism considering system performance reliability

“…When the system has only one failure mode, the cost for improving reliability linearly decreases with the rating scale, and as the ranking level increases, the cost for improving reliability will decrease. When the subsystem has multiple failure modes, similar conclusions can be obtained from formulas (23) and (25).…”

Reliability Allocation for Series‐Parallel Systems subject to Potential Propagated Failures

“…However, if such a relationship is not available especially in the early design stage, the system reliability requirement can be proportionally assigned to the corresponding subsystems or components according to their weights. Typical reliability allocation methods using the allocation weights include Bracha method, Aeronautical Radio Inc. (ARINC) method, Advisory Group on Reliability of Electronic Equipment (AGREE) method, the feasibility of objectives method, Karmiol method, the maximal entropy ordered weighted averaging method and the integrated factors method . Commonly used factors consist of complexity, failure frequency, state‐of‐the‐art technology, criticality, environmental conditions.…”

“…Typical reliability allocation methods using the allocation weights include Bracha method, 3 Aeronautical Radio Inc. (ARINC) method, 4 Advisory Group on Reliability of Electronic Equipment (AGREE) method, 5 the feasibility of objectives method, 6 Karmiol method, 7 the maximal entropy ordered weighted averaging method 8 and the integrated factors method. 9 Commonly used factors consist of complexity, failure frequency, state-of-the-art technology, criticality, environmental conditions. They are usually considered in different combinations to calculate the reliability allocation weight.…”

Footprint of knowledge acquisition improvement in failure diagnosis analysis

“…Furthermore, the parameters in conventional fuzzy FTA models cannot exactly signify the increasingly complex knowledge‐based systems. Secondly, due to the expert knowledge acquisition process, the domain expert board usually establishes diverse experience and knowledge from one another and produces various types of knowledge information including complete/incomplete, precise/imprecise, and known/unknown because of its cross functional and multidisciplinary nature . The latter shortages can be handled by the interval 2‐tuple linguistic terms method.…”

“…Secondly, due to the expert knowledge acquisition process, the domain expert board usually establishes diverse experience and knowledge from one another and produces various types of knowledge information including complete/incomplete, precise/imprecise, and known/unknown because of its cross functional and multidisciplinary nature. 27 The latter shortages can be handled by the interval 2-tuple linguistic terms method. The 2-tuple intuitionistic fuzzy numbers (IFNs) and many of its considerable extension can overcome the aforementioned limitation.…”