This paper addresses the question of how to support the designer with appropriate knowledge during conceptual design. It begins with a discussion of knowledge-based support for design and is followed by a scenario acc...This paper addresses the question of how to support the designer with appropriate knowledge during conceptual design. It begins with a discussion of knowledge-based support for design and is followed by a scenario account of the use of a Knowledge Support System. A system is described that demonstrates interaction with different forms of knowledge in concept vehicle design.It supports the creation of new designs by way of a solution generation and evaluation process that relies upon co-operation between the designer and the knowledge system. The results of user evaluation gave rise to a current research agenda which addresses the requirements of a multi-user platform for a design knowledge support environment for collaborative team work.展开更多
In order to compete in the global manufacturing mar ke t, agility is the only possible solution to response to the fragmented market se gments and frequently changed customer requirements. However, manufacturing agil ...In order to compete in the global manufacturing mar ke t, agility is the only possible solution to response to the fragmented market se gments and frequently changed customer requirements. However, manufacturing agil ity can only be attained through the deployment of knowledge. To embed knowledge into a CAD system to form a knowledge intensive CAD (KIC) system is one of way to enhance the design compatibility of a manufacturing company. The most difficu lt phase to develop a KIC system is to capitalize a huge amount of legacy data t o form a knowledge database. In the past, such capitalization process could only be done solely manually or semi-automatic. In this paper, a five step model fo r automatic design knowledge capitalization through the use of data mining is pr oposed whilst details of how to select, verify and performance benchmarking an a ppropriate data mining algorithm for a specific design task will also be discuss ed. A case study concerning the design of a plastic toaster casing was used as an illustration for the proposed methodology and it was found that the avera ge absolute error of the predictions for the most appropriate algorithm is withi n 17%.展开更多
文摘This paper addresses the question of how to support the designer with appropriate knowledge during conceptual design. It begins with a discussion of knowledge-based support for design and is followed by a scenario account of the use of a Knowledge Support System. A system is described that demonstrates interaction with different forms of knowledge in concept vehicle design.It supports the creation of new designs by way of a solution generation and evaluation process that relies upon co-operation between the designer and the knowledge system. The results of user evaluation gave rise to a current research agenda which addresses the requirements of a multi-user platform for a design knowledge support environment for collaborative team work.
文摘In order to compete in the global manufacturing mar ke t, agility is the only possible solution to response to the fragmented market se gments and frequently changed customer requirements. However, manufacturing agil ity can only be attained through the deployment of knowledge. To embed knowledge into a CAD system to form a knowledge intensive CAD (KIC) system is one of way to enhance the design compatibility of a manufacturing company. The most difficu lt phase to develop a KIC system is to capitalize a huge amount of legacy data t o form a knowledge database. In the past, such capitalization process could only be done solely manually or semi-automatic. In this paper, a five step model fo r automatic design knowledge capitalization through the use of data mining is pr oposed whilst details of how to select, verify and performance benchmarking an a ppropriate data mining algorithm for a specific design task will also be discuss ed. A case study concerning the design of a plastic toaster casing was used as an illustration for the proposed methodology and it was found that the avera ge absolute error of the predictions for the most appropriate algorithm is withi n 17%.
