Optimum Machine System Design Based on Re-Construction and Contraction of Design Spaces

摘要

A design optimization method of machine products based on flexible design spaces is proposed, in which optimum design solutions are searched for by breaking through the barriers of limited feasible design spaces, such as are engendered by preconceptions of designers, by effectively expand- ing and contracting the design spaces. This method has two steps: ( 1 ) going back to design states where higher degrees of freedom for the design are included, and ( 2 ) contracting expanded design spaces and searching for the optimum design solution using design sensitivity analyses and hybrid genetic algorithms. Design information is represented by networked frame structures that, during processing, expand the design spaces so that far more design possibilities are included. Hierarchical coding methods for solving expanded problems by Genetic Algorithms are constructed, and hybrid- type Genetic Algorithms optimization and local search methods are proposed. The foregoing methods are applied to design problems of mechanisms for transporting an object in a working space having obstacles.