Machine Identification Of Potential Assembly Process Failure Modes Based On Process Constituent Elements And Ontology

Assembly is a key link in the product manufacturing process.In order to effectively prevent quality problems in the assembly process,it is required to carry out process failure mode and effects analysis(PFMEA)on the assembly process.One of PFMEA’s important tasks is to identify all potential process failure modes.Faced with complex assembly activities,relying on manpower to exhaustively identify potential process failure modes is not only labor-intensive and time-consuming,but also lacks completeness in the recognition results.It has important theoretical research significance and practical application value to effectively solve the problem of automation of potential process failure mode identification and completeness of identification results in the assembly process.To this end,the paper proposes a machine identification method based on the potential constituent process failure modes of the process components and ontology.Firstly,the knowledge base of assembly process ontology is constructed.One is to construct the knowledge base of assembly process ontology based on process components.The assembly process knowledge processing step to work for the smallest unit assembly process,according to the content of the assembly process,to extract the content of can fully and completely describe the process elements,the step from the input,value-added processing activities,resources,environment,quality inspection and control,output six factors to describe,implements the assembly process activities described theoretically completeness.The second is to build a knowledge base of assembly process ontology based on semantic analysis.The assembly process ontology library based on semantic analysis is constructed by introducing natural language processing technology and processing process knowledge with semantic analysis.Thirdly,the two are integrated together to comprehensively represent the assembly process knowledge information and form a comprehensive knowledge base of assembly process ontology.The example set is mainly the rudder steering gear installation,soft fuel tank installation and landing gear installation of a certain aircraft.Secondly,build a rule base for generating potential failure modes of assembly processes.By analyzing the semantic structural characteristics of each process component and its process failure criterion in engineering practice,the use of semantic opposition analysis and its complementary technology and the string built-in functions in SWRL rules to establish a potential process failure mode based on process components generate rules.In view of the fact that the process components are extracted based on the work step,and the final potential process failure mode is identified based on the work step,a transfer rule is established to convert the potential process failure mode of the process component to the potential process failure mode under the corresponding process step or generate rules.The third is to use the Drools inference engine based on the aforementioned assembly process ontology knowledge base and potential assembly process failure mode generation rule base to perform inference recognition of potential process failure modes.In order to facilitate the statistical and analysis of the identified potential process failure modes,SPARQL technology is used,and the identification results of the rudder installation,soft fuel tank installation and landing gear installation of a certain type of aircraft in the assembly process knowledge base are evaluated and analyzed.The application research results of the paper show that the proposed machine identification method based on the potential assembly process failure mode of the process components and ontology has a good recognition effect,not only can quickly infer the potential process failure mode of the working step,but also high recognition accuracy.The ontology knowledge base and rule base constructed in this paper are both extensible.The ontology knowledge base model can also be used in other fields.The generation rules of potential process failure modes can also provide reference for the formulation of other rules.