Research On Process Knowledge Acquisition Method And System Design

The diversified and individualized needs of products have caused manufacturing companies to gradually shift to a multi-variety,small-batch production mode,which has led to a sharp increase in the demand for process knowledge of related parts.At the same time,the informationization and digitalization of enterprises have gradually improved and a large number of process data have been accumulated,but they cannot be fully utilized,showing the phenomenon of "rich data,lack of knowledge".This article takes process data as the research object,and excavates three aspects of part image,process route and process association rules to achieve process knowledge acquisitionThis article firstly,the image retrieval method based on SIFT features is proposed to obtain similar parts,and the improved Canopy Kmeans clustering algorithm is used to obtain the word frequency vector.The word frequency vector is used to retrieve similar part drawings.The SIFT feature has the feature of constant scale,which can effectively retrieve similar part drawings with different proportions,angles and positions,and has higher inspectionThen,mining the process route and process content respectively.In the process route-based rule mining,considering the complexity of process expression,a data preprocessing method for word segmentation and automatic data conversion is proposed.Then AP clustering algorithm without specifying the number of clusters is used to obtain typical process route,and PrefixSpan sequential pattern mining algorithm is used to obtain process route.For mining process content,an Apriori algorithm based on hash technology is proposed to improve the efficiency of Apriori algorithm.The validity of the three mining algorithms is verified by testing the process data.Finally,a process knowledge mining web system is designed and implemented based on the process knowledge mining method.In order to ensure the completeness of process data,a process data warehouse was built using integrated data warehouse technology.On the basis of clarifying the functional requirements of the system,based on architecture of B/S,the detailed design of the system process knowledge mining,data extraction,and system setting function modules was completed.And the main functions of the system are tested to verify the stability and feasibility of the process knowledge mining system.