A state reduction approach to interpretation of system behavior based on system simulation

System modeling and simulation techniques have been introduced to the early stage of model-based design in product development. Given a description of a system in terms of system components and connections, these techniques compute its numerical behavior by building and solving an equation system derived from the topology of system components. Currently, these techniques have a limitation in supporting designers to interpret the overall behavior of a system in terms of semantics rather than numerical values. To overcome the limitation, the paper briefly introduces a semi-automated procedure to support designers to validate a product model with system simulation results. The procedure reduces the dynamic state of a system by clustering the behavior of system components in terms of their parameters at discrete time intervals. It also generates the overall system behavior by computing clustering patterns and its transitions and identity relations across intervals in the entire simulation duration.