Research And Application Of Extension Association Rule In Product Conceptual Design

Conceptual design is an important stage of product design, and intelligent conceptual design isthe hot spot of conceptual design research. Extenics is a formal, logical and mathematical subject forthe intelligent treatment of contradictory problems. It provides a formal, modeling, qualitative andquantitative method to research conceptual design. For the features of innovation, diversity, hierarchyand so on of product conceptual design, from the viewpoint of intelligent design and knowledge reuse,this paper researched the application of product information extension modeling, extensionassociation rule mining and extension association rule reasoning in conceptual design by introducingextension theory and method, and applied the theoretical research to practical product design.First, for the deficiency of exsiting knowledge representation methods, the basic-element theoryof extenics is introduced to research the product information modeling. The different kind of designknowledge in conceptual design is formally described by basic-element. Extension knowledgerepresentation methods are studied by combining the basic-element theory with classical knowledgerepresentation methods, then the qualitative and quantitative product information basic-element modelis established; Next, for the design knowledge and experience implicated in the conceptual designprocess, data mining is combined with extension theory, an improved FP-Growth algorithm ofextension association rule mining is proposed, basic-element trancaction set is mined for associationrules based on the correlation and implication of basic-element, the association rules are extendedbased on the extensibility of basic-element and extension transformation, the knowledge acquisitionmethods is enriched, and the feasibility and efficiency of the method is verified by an example; Then,for the limitation of traditional reasoning method when dealing with fuzzy and uncertainly knowledge,an extension association rule-based fuzzy reasoning mothed is studied. The matching features ofdesign requirement and association rules are studied, a matching algorithm of extension associationrule is given based on the dependent function, the traditional criterion formula of relational degree isimproved, the optimization algorithm of extension association rule is studied, and the feasibilityand efficiency of the method is verified by an example; Finally, based on the theoretical studies above,a hydraulic turbine type-selection scheme design system is developed, then computer aidedconceptual design is realized based on extension association rule mining and reasoning.