一种基于模糊数算术运算的可靠性分析方法

可靠性是产品在规定的条件下和规定的时间内完成规定功能的能力。传统的可靠性分析方法是基于概率论的，其失效概率通常应根据大量的第一手数据进行估计。然而，实际上不可能收集到足够多的第一手数据。因此，这些失效概率通常是通过其它领域的参考数据，根据专家的工程和经验来判断估计的，这类判断通常与主观性和模糊性密切相关。为了处理这种主观性和模糊性，需要一种基于模糊数学的可靠性分析方法。提出了一种基于三角形模糊数算术运算的可靠性分析方法，并给出了应用该方法的一些例子。结果表明，该方法可以较好地表示失效率数据中存在的不确定性，而且其计算也十分简单。

A Reliability Analysis Approach Base on Fuzzy Number Arithmetic Operations

Reliability is the product's ability to perform its specified functions under the stated conditions for a given period of time. The conventional reliability analysis methods are based on the probability theory. Usually, the failure probabilities should be estimated from a large amount of data. In practice, however, it is not likely to collect enough immediate data. Therefore, the failure probabilities are usually estimated from reference data in other areas by experts based on their engineering and experienced judgement. The judgement is closely related to the subjectivity and fuzziness. In order to deal with this kind of subjectivity and fuzziness, a reliability analysis method based on fuzzy mathematics is necessary. In this paper, a reliability analysis approach based on triangular fuzzy number arithmetic operations is proposed, and some examples using this approach are given. The results show that the proposed approach can represent the uncertainties in the failure rate data fairly well with the advantage of a great computational simplicity.