Complex system maintainability verification with limited samples

Abstract Complex system maintainability verification is always a challenging problem due to limited sample sizes. Consequently, conducting maintenance experiments in a laboratory environment is an appropriate way to obtain data for maintainability verification. In maintenance experiments, faults are seeded in the equipment and maintenance activities are implemented to record repair time. In this process, two problems arise when laboratory experimental data (in-lab data) are used together with field data during the operational test and evaluation stage. The first problem is the verification of segmental maintenance data and the second one is the combination of in-lab data and field data for integrative maintainability verification. Regarding the problems mentioned above, this paper proposes a suitable methodology to solve them. Firstly, the idea of segmentally weighted verification is adopted and the segmentally weighted verification (SWV) method is proposed to realize in-lab data verification. Secondly, the Dempster–Shafer (D–S) evidence theory based integrative verification method is presented to solve the problem of in-lab and field data combination. A case study concerning radar system maintainability verification is presented as an example of the implementation of complex system maintainability verification in industry.

[1]  Qiang Miao,et al.  D-S evidence theory based maintenance evaluation under the situation of limited samples , 2010, 2010 Prognostics and System Health Management Conference.

[2]  Massimiliano Giorgio,et al.  Assessing high reliability via Bayesian approach and accelerated tests , 2002, Reliab. Eng. Syst. Saf..

[3]  Kai Goebel,et al.  Sensor validation and fusion for gas turbine vibration monitoring , 2003, SPIE Defense + Commercial Sensing.

[4]  H. F. Durrant-White.  Consistent integration and propagation of disparate sensor observations , 1987 .

[5]  Xavier Briottet,et al.  Presentation and description of two classification methods using data fusion based on sensor management , 2001, Inf. Fusion.

[6]  E. C. Goulden.  An analytic approach to performing a maintainability demonstration , 1990 .

[7]  Julia V. Bukowski,et al.  Bayesian reliability evaluation of computer systems , 1984 .

[8]  Djamil Aïssani,et al.  Impact of few failure data on the opportunistic replacement policy for multi-component systems , 2010, Reliab. Eng. Syst. Saf..

[9]  Arthur Mynett,et al.  The construction of confidence intervals for frequency analysis using resampling techniques , 2004 .

[10]  Juan Carlos Figueroa García,et al.  A Fuzzy Logic Approach to Test Statistical Hypothesis on Means , 2008, ICIC.

[11]  H. Albrecher,et al.  Asymptotic Results for the Sum of Dependent Non-identically Distributed Random Variables , 2009 .

[12]  James Llinas,et al.  An introduction to multisensor data fusion , 1997, Proc. IEEE.

[13]  Hugh F. Durrant-Whyte,et al.  Consistent Integration and Propagation of Disparate Sensor Observations , 1986, Proceedings. 1986 IEEE International Conference on Robotics and Automation.

[14]  Enrico Zio,et al.  Building confidence in the reliability assessment of thermal-hydraulic passive systems , 2009, Reliab. Eng. Syst. Saf..

[15]  Xianfeng Fan,et al.  Fault diagnosis of machines based on D-S evidence theory. Part 1: D-S evidence theory and its improvement , 2006, Pattern Recognit. Lett..