Case-based reasoning approach to concurrent design of low power transformers

Abstract In current practice, product design and process design of low power transformers are done sequentially in which quite a number of activities are mainly performed by engineers. Therefore, the time for the development of low power transformers and the quality of the product and process design are largely dependent on the experience of the engineers. In this paper, an artificial intelligence technique, case-based reasoning, is introduced to perform the concurrent product and process design of low power transformers from which a prototype concurrent design system for low power transformers, named CBS-TX, was developed. The prototype system aims to assist design engineers in formulating a configurable design of low power transformers in short time and provide the process information in the early design stage. In the system, the previous successful design cases and their corresponding process solutions are stored in a case library. After the input of design problems, the case-based system performs the case retrieval, and similarity analysis. A number of the closest matching cases are then selected and adapted. After the adaptation, case evaluation is then performed on them based on techniques for order preference by similarity of ideal solution (TOPSIS) algorithm and hence a concurrent design solution of low power transformers can be obtained.