The Logical Organization And Query Of Pdm Data Based On NoSQL Technology

In the large manufacturing industry, the product data stored in the PDM (Product DataManagement) becomes more and more, and shows the characteristics of “big data”, due to itsmany years accumulation, the rising number product model, and the implementation of the“configuration” mechanism. With the implementation of the MBD/MBE (Model BasedDefinition, Model Based Enterprise) technology and product whole life cycle management,not only make structure features of the engineering data become more and more weak, butalso make the users almost “all staff participation” in the Enterprise. These characteristicsabove have resulted in the increasingly serious problems in the aspects of high concurrencyreading and writing, high efficiency storage and access in mass data, high scalability andavailability of the PDM system based on RDB, such as more high-grade server and userterminal, larger number and capacity of the storage equipment, slower speed of user accessand longer data backup time. So the research of the new type storage and query of data inPDM has significant implications.Firstly, this paper analyzes the research status at home and abroad, and the concepts andtheories related. It analyzes the core content of the product structure and configurationmanagement in PDM—the creation concept of the product structure tree, and combine withstructure relations in product structure tree and the characteristics of the NoSQL. It discussesthe data storage structure of PDM data in NoSQL—MongoDB, which breaks the traditionalmodel of table structure in relational database. We establish the new model of data storage inPDM based on NoSQL.Secondly, the advantages and disadvantages of the MapReduce model are analyzed. Andaiming at the time-consuming problem of extracting single product structure tree from massproduct structure data, the strategy of MapReduce based parallel distributed computing modelis proposed to query all product structure tree of the mass product data.Finally, experiments are conducted to test and verify the strategies proposed in this paper.The experimental results show that the product structure data organization of PDM based on NoSQL can show the product structure relationship more clearly and maintain the productstructure tree more easily, and the MapReduce model based method of querying all productstructure tree from mass product data is more efficient than the traditional method.