Research On Knowledge Management Of The Manufacturing Process

Knowledge Economy is coming now. Like capital, labor and material, knowledge can be used as a factor of production. In the new century, the competition among the enterprises has transformed into the competition of knowledge they owned. In the new era that winning depends on knowledge, enterprise has to manage knowledge in order to maintain its competitive power.In the manufacturing process, almost all the sections such as R&D., Maintenance of Equipment Service after Sale and Condition Monitoring have the requirement to execute knowledge management (KM).Presently, KM is still in the primary phase in the application of the manufacturing process. The reseach work is mainly based on technologies that supporting knowledge sharing and reusing. During carrying out KM in the manufacturing process, the author proposes that managing and using existing knowledge through knowledge sharing is only the base of the whole activity. If the enterprise wants to maintain its core competition power, it has to innovate.The research of this dissertation is focused on the knowledge management especially knowledge innovation of the manufacturing process. The main contents of this dissertation are as follows:(1) After analyzing the requirement of KM of the manufacturing process systematically, the author not only proposes the framework and model ofKM of the manufacturing process, but also explores how to combine it with the traditional methods used in process such as ERP> PDM^ SPC and Condition Monitoring, etc. According to the practical requirements of manufacturing process, the paper proposes a loose-tight knowledge management model to combine knowledge innovation and knowledge sharing. The author not only designs the environment model (A-Dynasites Model) needed by the KM activity, but also introduces knowledge evolution model (SER Model) to help promote the environment.(2) In the manufacturing process, the collected data always has the feature of fuzzy, incomplete, noisy, etc. So it is important to explore how to catch the character of the measured object through analyzing these data. Combining with the designing and using of apparatus to measure the profile of roller, in the data preprocessing aspect, the author proposes a new filtering method — General Regression Neural Network (GRNN). It not only has the robust feature but also is more efficient than the traditional method. In the research of data fusing technology especially error separation technology (EST), the author not only analyzes how to use the EST with the noisy data, but also proposes a new EST method — Statistically Second-Point Error Separation Method.(3) Knowledge discovering is the core of knowledge innovation. After getting the true character of the measured object, the author analyzes how to discover the potential useful knowledge from the data. The author pays more attention on one of the knowledge discovering methods — association rule. According to the requirements of manufacturing process, after combining with many technologies and methods such as constraint technology> closed itesmets theo^ concept lattice> inference technology and estimation technology, the author proposes a series of new algorithms. The efficiency of them has been proved by experiments.