| A generalized evaluation methodology for the acceptability of mechanism configurations performing a variety of motion synthesis programs is addressed in this thesis. This evaluation methodology includes: constraints on number of synthesis positions and geometric structure analysis. Using "deep knowledge representation" and "deep reasoning", a knowledge based mechanism configuration evaluation system, named "KBMCES", is developed on an IBM PC computer. The system uses frames and rules for knowledge representation. A constraint based reasoning strategy is applied to its inference engine. The system has been successfully applied to evaluate planar mechanisms with one degree freedom, up to eight links, and revolute joints. This study will allow one to fully understand geometric structure of planar mechanisms and potential applications of planar mechanisms. Moreover, this evaluation methodology will provide a new approach for intelligent mechanism configuration selection and creative mechanism design. |
