Computer aided process planning based on ontology and rules system applied to forging domain
In order to reduce product and process plan design time and cost, the integration of the relations that exist between product data and manufacturing data is essential. This paper proposes an original approach to support this integration, mix of rule based and classification approaches with ontologies. This approach has been implemented in a computer tool which allows the formalization and the use of manufacturing knowledge, particularly to aid the design of process plan in the forging manufacturing domain.
| Attachment | Size |
|---|---|
| IPROMSv1.wmv | 7.34 MB |
The use of conditions is done with forward chaining: consequences on the process plan are infered starting from the data of the part. The goal isn’t to get presumed final conclusion(s)/solution(s), but just identify the branches of the process plan scheme that are not possible (because conditions on these branches are not respected), so we don’t use backward chaining.
The transformations that are children of a XOR node cannot be in a same process plan deducted from the process plan scheme. This kind of node permits us to agregate several process plans in a same process plan scheme.
Dear authors,
Have you published some of the results obtained using your system? How do the process plans produced by the system compare with plans generated by people?
Thank you.
D Pham.
My Questions :
1. What kinds of searching process method that you used in rule based transformation where you can get the final conclusion?, complete it with the reasons.
2. Why do you use XOR as your operator when representation your rule based?
Thank You.