Constraint Satisfaction Method for Networks of Design Variables and Constraints in Mechanical CAD.
Joo-Heon Cha, Masaaki Yokoyama, Itaru Okabe
Abstract
Open-access reader
Joo-Heon Cha, Masaaki Yokoyama, Itaru Okabe
Abstract
Open-access reader
We describe a design process by using constraint networks which consist of constraint nodes and attribute nodes. The constraint nodes contain design formulae, and the attribute nodes represent design variables. The constraint representation allows the system to describe an object declaratively. However, as the constraint programming doesn't give a structural description of the design process, it is difficult to apply the constraint programming to a bigger system. In order to improve this difficulty, we present a new method that divides a complex constraint network into several constraint modules in which every module is composed of a simpler and smaller constraint network. Moreover, we present a new method to solve the constraint network problems with closed loops. The usefulness of these methods is illustrated by application to the design of a gear system.
OpenAlex reports 1 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
We describe a design process by using constraint networks which consist of constraint nodes and attribute nodes. The constraint nodes contain design formulae, and the attribute nodes represent design variables. The constraint representation allows the system to describe an object declaratively. However, as the constraint programming doesn't give a structural description of the design process, it is difficult to apply the constraint programming to a bigger system. In order to improve this difficulty, we present a new method that divides a complex constraint network into several constraint modules in which every module is composed of a simpler and smaller constraint network. Moreover, we present a new method to solve the constraint network problems with closed loops. The usefulness of these methods is illustrated by application to the design of a gear system.
Key concepts: Constraint logic programming, Constraint (computer-aided design), Constraint satisfaction, Constraint programming, Binary constraint, Constraint graph, Local consistency, Constraint satisfaction dual problem