Knowledge Induction for Injection Mould Repairs Based on Rough Set

The procedures used in previous injection mould repair schemes are viewed as valuable knowledge by many mould manufacturers. This knowledge is stored in computer on a case-by-case basis and can be used to form new repair schemes when needed by searching the database. With the enlargement of the database as more case is added, the efficient retrieval of the correct injection mould repair scheme has relied on the knowledge and skill of the operator. Based on the analysis of the representative characteristics of each injection mould repair scheme, knowledge induction model based on rough set for repair schemes is firstly put forward in this paper. As the basis of the model, feature and concept hierarchy model (FCHM) is presented from two points of view which are feature hierarchy and concept hierarchy. Then, knowledge representation for repair schemes can be provided by using FCHM. After reconstruction and fuzzy process for repair schemes, knowledge induction can be done by using attribute reduction and rule induction based on rough set. Finally, through the toolkit software of rough set theory named ROSETTA, an experiment is carried out to induce knowledge from repair schemes in accordance with the representative models on different layers.