An integrated knowledge-based system for sheet metal processing
This paper presents an integrated knowledge-based CAD/CAM system for sheet metal cutting-punching combination processing. The whole system consists of five functional modules, i.e. modeling, nesting, process-planning, NC-programming and simulation & reporting module, which are integrated seamlessly in terms of data processing. The knowledge base is the core of the integrated system, based on which, the system runs in an efficient and intelligent manner. This system has been successfully applied in industry, and an application example is demonstrated.
The slides of your presentation are not in the proper order, for example the case study and the conclusion slides come before the introduction, modular structure and knowledge base slides. You can easily adjust this problem by setting up the field "weight"at the bottom of the page when you edit each of your slides. In addition, it would be beneficial for your presentation if you could include some speaker notes for each slide. This can be done when editing a slide by writting such notes in the field "excerpt".
Emmanuel Brousseau
Thank you very much for your suggestions. The slides have been re-ordered and relevant notes are included.
At present, only the DXF format is considered in our case. Actually, in our practical applications, geometrical modeling is not a big problem, DXF is enough for the use. The key point is the KB. Of course, a neutral standard format should be helpful for the system's universality, which will be considered for the further research.
Knowledge acuisition is probably the most time-consuming task in developing knowledge-based systems. Automatic knowledge acquisition techniques such as inductive learning, insatance-based learning and other machine learning techniques have been developed to address this problem. Have you considered using any of these learning techniques?
What programming language/environment do you use?
Since the knowledge in NRB/PRB is essentially empirical, the KB of the system is basically built up according to the experiences of the technicians in the plants. Now we are trying to use the case-based learning technique for the expansion of the KB.
Is there any limitation on thickness of the stock-sheet to apply PRB?
Thank you for your timely answers to all the comments to your paper. You mentioned that the puncher moving distance can be formulated as a "traveling salesman model" and that you adopted the nearest neighbour algorithm to solve this problem. As you also pointed out, a lot of algorithms have been developed to address the traveling salesman problem. Is there any particular reasons for you to choose the nearest neighbour algorithm?
Emmanuel Brousseau
For the sequencing of punching process, the main objective is to minimize the frequency of puncher change rather than the total puncher-moving time.The NNA is simple and efficient enough for our problem, which is basically samll/medium sized.
Thank you for your contribution to IPROMS2005. You mention in your paper that the geometrical information processed by the modeling module of your system is converted from outside CAD softwares. In this case, via a DXF interface from AutoCAD. Did you consider using a neutral format, like STEP Application Protocol 203 for example, as an input to this module? Such a neutral fromat can be output by most commercially available CAD systems and this could be an advantage over the DXF format.
Emmanuel Brousseau
The Manufacturing Engineering Centre
Cardiff University
Cardiff, CF24 3AA