Research On Intelligent Recommendation And Review Technology For Assembling Process Of Aerospace Engine Core Components

With the increasing of aerospace engine’s models and types,the design and assembly of aerospace engine have also show diversity.There are many kinds of assembly processes for engine core components,but the reusability of historical processes is low.At the same time,the process design has greater experience dependence,which leads to long cycle and low efficiency of process design.In addition,the process design specifications are not uniform and problems such as incorrect parameter setting and nonstandard process description are likely to occur.Therefore,it is of great significance to research the intelligent recommendation and review technology of engine core components’ assembly processes for improving process design efficiency and quality.This paper will study mining technology of sequential association rules to improve reuse rate of historical process plans,and establish the intelligent recommendation model of engine core components’ assembly processes to improve the efficiency of process design.Finally,this paper will also study the engine core components’ assembly process review rule representation method and review algorithm and complete the procedure of process document review to improve the quality and accuracy of process design.The specific research methods are as follows:First of all,in order to improve the reusability of aerospace engine core components’ assembly processes and resolve the shortage that the traditional association rules mining algorithm is difficult to realize sequential mining,this paper will study the sequential association rules mining technology to realize the mining of typical process sequences in historical assembly process plans.Based on the analysis of exited association rule mining methods,a sequence based FPGrowth algorithm based on conditional confidence is proposed to realize the mining of sequential association rules.Comparison experiments with Apriori algorithm and FP-Growth algorithm show that the algorithm is effective and then the typical process sequences of historical assembly process are excavated.Then,in order to improve the efficiency and reduce the experience dependence of process design,this paper will study the intelligent recommendation method of aerospace engine core components’ assembly processes.Based on the historical assembly process content,the process knowledge graph is constructed to establish the relationship between the each part of the process file;By using the sub graph retrieval function of knowledge graph and combining the assembly process association rules,the assembly process recommendation model based on Noisy-OR＿BN will be constructed,and then,the process design recommendation list is obtained based on Bayesian network reasoning ability.The effectiveness of the process recommendation model will be verified by the historical process plan.Next,in order to improve the standardization,accuracy and design quality of process documents,this paper studies the rule-based process review algorithm.Based on the analysis of knowledge representation method and the characteristics and requirements of process review,the data review rules of keywords and assembly process are standardized to create a process review rule base.And the rule-based process review algorithm is studied to review the standardization and accuracy of process documents.Finally,based on the research of the above theoretical methods and combined with the requirements of aerospacce engine core components’ assembly process design,the assembly process management system is developed to complete the functions of assembly process data management,process plan design and process review,which provides system support for actual assembly process design and review and effectively improves the efficiency and quality of assembly process design for core components of aerospace engine.