Affiliations: [a] General Electric Manufacturing Company, Hong Kong | [b] The Hong Kong Polytechnic University, Department of Industrial and Systems Engineering, Hong Kong | [c] The University of Warwick, Dept. of Engineering, Warwick Manufacturing Group, UK
Abstract: Use of Computer Aided Engineering (CAE) for design validation in a product development cycle has been pursued as a state-of-the-art strategy to compete in the global competition. However, the combined use of CAD and CAE is still impractical to solve non-linear design problems that are multi-disciplinary in nature, even allowing for the speed offered by the advancement of computer computation. In order to bridge the gap for the full deployment of a computer aided/virtual design alternative, a methodology that involves the combined use of CAE and Artificial intelligent (AI)/knowledge intensive approach is proposed. Starting with a breakdown of attributes of a design task, a number of potential artificial intelligent algorithms are screened. Followed by a subsequent training schema, the most suitable AI algorithm is then selected. After the crystallization of the final knowledge database, the application can then be deployed into modules for subsequent or further new applications. The proposed model was illustrated with a case study for the design of a plastic toaster cover with respect to a heat test.
