Font Size: a A A

The Study On Bucket Elevator Designing Expert System Based On CBR And RBR

Posted on:2007-01-15Degree:MasterType:Thesis
Country:ChinaCandidate:F WangFull Text:PDF
GTID:2132360182483054Subject:Mechanical Manufacturing and Automation
Abstract/Summary:PDF Full Text Request
One of Artificial Intelligence research fields is expert system technology. Inthis dissertation, Case-based reasoning technology and traditional CADtechnology are applied to the bucket elevator designing process, and a bucketelevator designing prototype expert system is built.Based on the basic theory of expert system and software engineering, aframework of the bucket elevator designing system is presented after theanalysis of the bucket elevator family structure. Also the working principle andworking process are discussed.As the case database building technology is the key technology in CBR, theprinciples and approaches to build the database is discussed here. First, theknowledge organizing strategy is studied, then, hiberarchy structure is broughtforward. The relational models for case representation and database structure aregiven. Case database is built with ACCESS, and principal axis subassemblycases' 3D parameter models are established with PRO/E.As the expert system solve problems based on the case and rule database,the approaches for developing the inference engine is studied. The mainparameters set which affects the structure is decided after the analysis of the allthe parameters. The fuzzy is also used to calculate similarity between case andnew project, and hereby the most similar case can be chosen.Using existing successful product cases and knowledge for the design ofbucket elevator can make the designing process with more automation and moreintelligence. In addition to that, the developed prototype system structure, casedatabase building and reasoning method can be used as a reference to developdesigning systems for other products.
Keywords/Search Tags:Expert System, Bucket elevator, Case database, Case-based Reasoning, Similarity measurement, Characteristic Modeling.
PDF Full Text Request
Related items