Research On Type-selection Optimization For Truss Structures Based On Expert System

Scheme design is the most creative phase of structural design, which needs plentiful experience and tight cooperation of architects and structure engineers due to its features of soft scientific decision making. The generation, development and widely application of expert system give a new tool to research structural optimization and offer a new method to resolve the sophisticated problems of structural optimization.The design process of truss structural form selection includes three main stages, which are scheme generation, scheme evaluation and scheme decision-making. The structural optimization knowledge of different truss types is collected through optimum design, and knowledge base of the expert system is built by series rules. Four objective factors that affect form selection are proposed in this article. The model of factor system influencing the form selection of truss structures is constructed according on the theory of data structure, basing on internal relations and logical relationship of sub-factors of every objective factor. Scheme evaluation is implemented by the multi-level and multiple-factor fuzzy comprehensive evaluation method. Fuzzy multiple attribute decision-making progress based on relative approach degree for selection is presented to decide the best scheme.The expert system for optional design of truss structure is built by using Visual Basic 6.0 as development tool of interface and driver, CLIPS as inside inference engine, and using Microsoft Access 2003 to set up data-base. The knowledge of the expert system is expressed by facts and production rules, and the inference mode is forward by using CLIPS. The system contains application, data maintenance, system maintenance and user query, having friendly interface. The knowledge base and data base arc easy to modify, expand and maintain.