The Data Organization And Management Of BOM Based On NoSQL

Bill of materials(BOM) is the core foundation data of manufacturing enterprises, in product design, simulation, manufacturing, maintenance and other life-cycle management process, which plays an extremely important role. On the one hand, the existing disorganized BOM management is due to the wide variety of BOM, the non-uniform data source and the wide range of departments, which seriously affected the development of enterprise products. On the other hand, with the use of model based definition(MBD) and demand of multi model parallel development, these make BOM the unstructured data, numbers of versions, complex hierarchies, and gradually big data states. These cause high concurrent read and write, massive data storage and access, high scalability and high availability become increasingly prominent in the BOM management system based on relational database(RDB). Therefore, it has important theoretical significance to research the BOM data organization and management technology.This thesis focuses on research on the data organization and management of BOM based on NoSQL. In the premise of meeting the massive BOM data storage, the data organization and traversal of BOM is realized with the help of NoSQL, which can avoid the problems such as BOM data redundancy, poor data consistency, low query efficiency and poor concurrency. The main objective of this thesis is to build a high concurrency, massive data storage, high available BOM prototype system based on NoSQL in enterprise private cloud platform.Firstly, BOM logical data organization presents network structure in general by analyzing the structure of BOM data and the logical relations between those data in PDM. With summing up the lack of BOM based on relational database, the NoSQL data organization of a single data source of BOM in PDM is proposed, which combines with the advantages of NoSQL and the characteristics of BOM data.Secondly, the traditional BOM traversal query algorithm has low efficiency and low efficiency. To address this problem, this thesis analyzes the query characteristics of BOM, and proposes a BOM traversal query strategy based on MapReduce to realize the fast query processing of BOM. This strategy takes the NoSQL database as the underlying data storage system and uses the MapReduce technology.Finally, this thesis constructs a BOM prototype system based on NoSQL in enterprise private cloud platform, and conduct extensive experiments to evaluate the performance of the proposed NoSQL data organization strategy of single BOM data source and BOM traversal query algorithm based on MapReduce in PDM. The results show that the system works well, and the method is effective and feasible.