Research on an exponential distribution reliability function model based on multi-source heterogeneous data supplements

Abstract

Purpose

As productions show characteristics of multi-varieties and small batch in a recent new product system, it is more difficult to acquire its failure rate data. With the help of expert experience information, the authors can get the interval estimation of failure rate data under different methods, so how to make the interval convergence with the new information is an important problem to be solved. The paper aims to discuss this issue.

Design/methodology/approach

In this paper, the concept of generalized standard grey number is used to characterize the multi-source heterogeneous uncertainty failure rate data into a unified framework. Then, the engineering construction method is used to calculate the average failure rate and build the grey exponential distribution reliability function, whose image is presented as the possible region of the two-curve envelope.

Findings

Further, according to the normal distribution assumption of the regional convergence based on the information supplement, the convergence problem of the reliability function is transformed into the convergence of the area of the curve envelope region, and construct the multi-objective programming model with the minimum envelope area and the lowest total cost of information acquisition, acquire the conclusion that the failure rate is equal to the nuclear of the average failure rate when the envelope region converges.

Originality/value

Through the case analysis of the equipment ejection system of the Harbinger system, five groups of results are obtained by Matlab simulation, which verify the rationality and feasibility of the model described in this paper.

Citation

Shao, H., Fang, Z., Zhang, Q., Hu, Q., Cai, J. and Tao, L. (2017), "Research on an exponential distribution reliability function model based on multi-source heterogeneous data supplements", Grey Systems: Theory and Application, Vol. 7 No. 3, pp. 329-342. https://doi.org/10.1108/GS-06-2017-0017