基于梯形区间二型模糊的失效模式、影响及危害性分析方法

发布时间：2018-05-07 04:23

本文选题：失效模式 + 影响及危害性分析　；参考：《昆明理工大学》2017年硕士论文

【摘要】：潜在失效模式、影响及危害性分析(Failure Modes,Effects and Criticality Analysis,FMECA)方法是一项在出售产品给客户之前,识别、确认和消除在系统、设计、过程和服务中存在的已知的和潜在的失效、故障、错误、问题等的工程技术。该技术应用于系统、设计、过程和服务的初期,利用现有的部分有效数据、历史相关文件信息以及来自各相关领域的专家所组成的FMECA小组成员的专业知识与经验,在风险发生前进行失效模式的查找与分析,进而造成了 FMECA方法特殊的应用环境——高度不确定的分析环境,其中包括:无法获得全部有效数据所造成的模糊不确定性;FMECA小组成员复杂多样的专业背景造成的专家成员之间的模糊不确定性;FMECA小组成员自身对失效模式判断的模糊不确定性。这种复杂的不确定性不可避免的造成了 FMECA方法量化分析部分的失真、失准问题。在此应用背景下,本文提出能够量化模糊不确定性的梯形区间二型模糊FMECA方法,提高了 FMECA方法量化分析的准确性和有效性。具体研究方法及主要研究成果有:(1)构建了信息的量化模型。首先,采取符合人类思考及判断能力的区间数作为初始个体判断信息的模糊不确定性量化数据;然后,通过对群体给出的这组区间数进行累积和堆叠,找到群体判断中隶属程度最高的判断区间作为群体判断在真值;同时兼顾除此之外的差异性数据,这部分数据信息记录了个体的不确定性以及个体之间的不确定性差异;最终构建出记录了群体判断共识与差异的梯形区间二型模糊群体决策值,并以该值作为进一步FMECA分析的风险因素评价数据。(2)提出了梯形区间二型模糊综合排序法。通过拆解梯形区间二型模糊的隶属函数从两部分分别进行比较:一部分是该模糊数所代表的真值部分,也就是模糊数隶属程度最高的部分,采用重心比较法进行比较;另一部分是该模糊数所记录的模糊不确定性部分,也就是模糊数隶属程度较低且较为分散的部分,采用基于相对几何空间大小及位置构建的针对隶属函数离散部分的比较法进行比较。综合两部分比较结果,并记录在不同数量级上,使得比较结果既保护了模糊数真值的有效性又充分考虑了模糊不确定性程度对该模糊数的影响。(3)提出基于梯形区间二型模糊的FMECA方法。通过构建的模糊信息量化模型,处理FMECA方法定性分析中的模糊不确定性并聚合成群体判断评价数据。再通过梯形区间二型模糊综合比较法,从两个部分对风险因素的群体评价值进行比较。最后,根据梯形区间二型模糊FMECA方法的风险优先系数(Risk Priority Number,RPN)计算值找到高风险、需优先解决的失效模式,并完成完整的梯形区间二型模糊FMECA方法。
[Abstract]:Potential failure mode, impact and hazard analysis FMECAA method is a method of identifying, confirming and eliminating known and potential failures, errors in systems, designs, processes and services before selling products to customers. Engineering techniques for problems, etc. The technology is applied in the early stages of systems, designs, processes and services, using existing partially valid data, historical documentation information, and the expertise and experience of FMECA team members from various related fields. The failure mode is searched and analyzed before the risk occurs, which leads to the special application environment of FMECA method, which is highly uncertain analysis environment. It includes: the fuzzy uncertainty caused by the failure to obtain all valid data and the fuzzy uncertainty among the expert members caused by the complicated and diverse professional background of FMECA members. This complex uncertainty inevitably leads to distortion and misalignment in the quantitative analysis of the FMECA method. In this context, a trapezoidal interval type fuzzy FMECA method, which can quantify fuzzy uncertainty, is proposed in this paper, which improves the accuracy and effectiveness of the quantitative analysis of the FMECA method. The specific research methods and main research results are: 1) the quantitative model of information is constructed. First, the fuzzy uncertain quantitative data of the initial individual judgment information is taken according to the interval number of human thinking and judgment ability, and then the group interval number is accumulated and stacked by the group. Find the highest degree of membership in group judgment as the true value of group judgment, and take into account the difference of the other data, this part of the data information records the uncertainty of individuals and the uncertainty difference between individuals. Finally, a trapezoidal interval type 2 fuzzy group decision value, which records the consensus and difference of the group judgment, is constructed and used as the risk factor evaluation data for further FMECA analysis. (2) the trapezoidal interval 2 fuzzy comprehensive ranking method is proposed. The membership function of trapezoidal interval type 2 fuzzy membership is compared from two parts: one part is the true value part represented by the fuzzy number, that is, the part with the highest membership degree of fuzzy number, and the center of gravity comparison method is used to compare; The other part is the fuzzy uncertainty part recorded by the fuzzy number, that is, the part with lower membership degree and more dispersion of the fuzzy number. A comparison method for discrete parts of membership function is proposed based on relative geometric space and position. Synthesizing the results of the two parts, and recording them on different orders of magnitude, The comparison results not only protect the validity of the true value of fuzzy number but also fully consider the influence of the degree of fuzzy uncertainty on the fuzzy number. (3) A FMECA method based on trapezoidal interval type 2 fuzzy method is proposed. The fuzzy information quantization model is constructed to deal with the fuzzy uncertainty in the qualitative analysis of FMECA method and to aggregate the evaluation data of group judgment. Then through trapezoidal interval type 2 fuzzy comprehensive comparison method, the group evaluation values of risk factors are compared from two parts. Finally, according to the risk Priority number FMECA calculated value of trapezoidal interval type 2 fuzzy FMECA method, the high risk and priority failure mode is found, and the complete trapezoidal interval type 2 fuzzy FMECA method is completed.
【学位授予单位】：昆明理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：F224;F273.2

【参考文献】