| The collaborative design and simulation technologies are essential in the product design process. The environment of the collaborative design and simulation facilitates the feedbacks and communications among different departments and improves the efficiency of the product development, but at the same time, it proposes higher requirements for the knowledge fields of the workers in related departments. They are expected to master some basic knowledge of other domains so that the communication and collaboration will be able to be carried out successfully. The advisor systems are based on the traditional expert systems, which can provide the related knowledge of different domains for the personnel in charge of the collaborative design and simulation and help the designing and evaluating. Moreover, by means of the advisor systems, the enterprise's knowledge can be accumulated and precipitated to train the new employees. Therefore the advisor system is an important AI toolset to support the multi-domain product design. Considering the features and requests of the collaborative design and simulation, this paper gives the architecture of a design and simulation advisor system; brings out a way of knowledge expression based on the improved production system, which uses the semantic mapping to realize the integration of the isomeric knowledge and facilitate the cooperation of them. The object-oriented technologies are borrowed to create the relationships among the modularized knowledge, which reduces the overlapping of the knowledge and facilitates its operations such as inserting and deleting. The knowledge base of the design and evaluation advisor system is divided into the captured experience base and the scientific principles base, so that the advisor system not only inherits the merits of expert systems but also becomes more reliable because of the added principles and scientific models. Based on the way of knowledge expression and the architecture of knowledge base, we search for the ways of how to use the data mining technologies to extract knowledge from the enterprise's history data base and how to store them into the knowledge base. Openness is a most important feature of the advisor systems, so the XML technologies are used to realize the dynamic facts' automatic adding. The method integrates the knowledge of collaborative design and simulation and improves the efficiency of utilizing the knowledge.On the basis of the theoretical research, a practical advisor system is realized for the design evaluation of the submarine's weapon electronic system. This system is running for test in Vessel System Engineering Department and it is estimated that the efficiency and intelligence of the design evaluation will be greatly improved so that the enterprise experts' experiences can be shared and solidified, also these experiences will be managed and maintained uniformly to promote the information management and integration of the enterprise. |