Based On The Complexity Of Semantic Case Base Reasoning Techniques

Complex Product represents single or batch produced products wit advanced technology and high R&D cost. They are usually included in multidisciplinary fields, such as mechanics, automation, electronics, software, hydraulics etc. The design and manufacture of complex product is a creative, complex, high-cost and high-risk long-term process, which necessitates the use of some assistant tools during this period. Case-based Reasoning (CBR) is one of the most important approaches frequently used in A. Its basic idea is to solve a new problem with experience obtained from previous problems. Accordingly, CBR may also guide the development of complex product if used and adjusted properly. However, the traditional CBR cannot meet the requirements in current complex product design due to its lack of semantic support and knowledge share ability. The Semantic Web technology is one of the hottest web technologies at present. It provides a universal framework with which we can share and reuse data across different programs, enterprises and entities. Applying the Semantic Web technology to CBR can thus solve the semantic support and knowledge share problem in the traditional CBR System.In this paper, the semantic complex product database and CBR were examined in order to construct the efficient semantic CBR System of complex product.Firstly, we analyzed the features and defects of the traditional CBR Systems. Then we brought forward an ontology and frame based representation Model (OFRM). OFRM solves the knowledge share and semantic support by describing the sub-case of the complex product in term of Ontology, and reflects the design idea of complex product (i.e., from simple to complex) by organizing the sub-cases into the whole complex case in term of frame.Secondly, the flow charts and policies of case retrieval as well as the similarity calculation methods were examined. Then we brought forward an improved semantic similarity calculation model (ISSCM) at the base of the Distance-based similarity calculation method. With OFRM and ISSCM, we constructed a semantic complex product case retrieval model (SCPEM) and a case reasoning model (SCPNM). Finally, based on the researches above, a sub-module in a special complex product Car was studied to testify the performance of the semantic complex case representation model (OFRM) and the semantic complex case retrieve model (SCPEN).