Optimized Retrieval for Manufacturing Resource Using Fuzzy Clustering under Uncertain Environment

摘要

Manufacturing Resource (MR) modeling, encapsulating, and retrieving are the basic and key supporting technologies. Retrieval methods based on text or semantic information suffer from the greater granularity, leading to lower retrieval accuracy rate, while methods based on exact numbers lack the ability to express the fuzziness and uncertainty existing in information on MR. Against the cases that the information on manufacturing resources often exhibits uncertainty and fuzziness, presented are the regulations of conversion of the information on MR to interval numbers and the corresponding normalization method. The hierarchical fuzzy clustering retrieval method is presented, first retrieving to cluster layer, and then to specific physics manufacturing resource layer. Finally, an example is illustrated to validate its effectiveness and practicability and the issues raised and the way to future research direction are discussed.