Study Of Computer Aided Assembly Sequence Planning Based On Graph Theory

The technology of CAD/CAM is very mature and widely used in production. As an important part of Computer Aided Engineering, Computer Aided Assembly Sequence Planning (CAASP) has important significance on improving product design, reducing cost and time to market. However, many problems concerning this issue remain unsolved. Therefore, it is especially important to conduct the research of Computer Aided Assembly Sequence Planning (CAASP) and its key technologies.Based on domestic and oversea research status and practical situation, this dissertation conducted intensive research on Computer Aided Assembly Sequence Planning (CAASP) and its key technologies. Moreover, a software "CAASPS" (Computer Aided Assembly Sequence Planning System) was developed. Major research contents and innovations are listed below:(1) Product assembly information model is the foundation of assembly sequence planning. With the consideration of automated assembly sequencing and information required for DFA (Design For Assembly), mathematical model was constructed to express assembly information in design stage of prototype and DCM (Dual-number Connection Matrix) was used to identify assembly precedence relationship among parts. This model is suitable for assembly planning without the availability of product CAD model.(2) Based on assembly information model, the idea of hierarchical assembly was applied in the generation of assembly sequence, by which the efficiency and practicality of assembly sequence planning were increased.(3) Assembly sequence planning algorithm on the basis of Graph theory was proposed in this study including the algorithms to detect possible sub assembly, verify feasible sub assembly and generate all available sequences. The big problem "combination explosion" raised from current research "Cut-Set" was successively solved.On the foundation of above algorithms, the software "CAASPS" (Computer Aided Assembly Sequence Planning System) was developed and verified with an actual assembly planning example. It is recognized that the algorithms are totally correct and effective.