A hybrid approach for the automation of functional decomposition in conceptual design

Abstract

ABSTRACTMultidisciplinary complex mechatronic systems span disciplines such as mechanics, electronics, control and software. Increasing complexity makes the implementation of automated design for complex mechatronics highly challenging. An effective method for decomposing an overall function to obtain the function structure and principle solution is needed for efficient creative design. A hybrid approach is proposed to automate the process of functional decomposition which contains three main steps. First, an overall function is decomposed to several functional effects automatically based on qualitative processing reasoning. Then the physical effect decomposition is conducted for each functional effect based on the support functions. For the functions for which a suitable physical effect cannot be found, a backward search decomposition is proposed to find a multichain function network for it. The decomposition process is implemented with SysML, and illustrated through an automobile engine example. An analysis is given to show the performance of the proposed method.