Research On Intelligent Potential Failure Mode And Effect Analysis

A research of Modern quality management shows that high-quality products are taken measures to prevent the occurrence of quality problems in advance,rather than improving quality through post-mortem testing.Therefore,it is essential for enterprises to consider the quality of the products at the beginning of the product design.Potential failure mode and effect analysis(FMEA)is a typical prevention reliability analysis method that evaluates reliability of product or system from the angle of potential failures and uses this information as a basis for the prevention and improvement of design planners.Therefore,accurate and efficient FMEA analysis has important significance for the enterprises to improve the product quality and the market competitiveness.This paper studies Intelligent FMEA from the aspects of FMEA knowledge modeling,knowledge retrieval and risk assessment.Firstly,through analyzing the features of FMEA knowledge,a new FMEA knowledge representation framework based on ontology is proposed.And taking FMEA object,function,failure analysis,FMEA team as four core concepts,analyzing the main concepts and their relationship of the four concepts,the corresponding ontology models are formed.The FMEA ontology model lays the foundation for subsequent FMEA knowledge retrieval and reuseSecondly,according to the shortcomings of the commonly used retrieval techniques,the semantics retrieval method of FMEA knowledge based on ontology is studied.And a semantic similarity calculation model considering semantic distance,semantic depth,semantic density,semantic coincidence degree and semantic attribute is proposed.Finally,a case study of the internal combustion engine piston is presented to verify the effectiveness of the proposed method.Thirdly,by analyzing the existing problems of the FMEA risk assessment method,an improved FMEA is proposed for considering fuzzy context and the relevance of the failure modes and failure causes.The method combines triangular fuzzy number,grey relational analysis(GRA)and the decision-making trial and evaluation laboratory(DEMATEL)method.Through considering the fuzzy of the risk assessment factors and the relationship of the failure modes and failure causes,this method makes the ranking result more consistent with the practical engineering.Finally,based on the above research,combining the product lifecycle management system——InforCenter PLM in China Aerospace Shandong Hoteam Software,we achieved the related functions of intelligent FMEA system.And the system is successfully applied in the FMEA analysis of a mechanical manufacturing enterprise.