Abstract
In systems engineering, determining a system's architecture is a crucial part of the preliminary system-definition phase because this architecture greatly impacts the quality of the system, the way the subsequent phases of the design process are organised and the overall performance of this process. Consequently, systems architects need methods for defining the couplings of modules (or sub-systems) and for estimating the impact of allocation decisions on candidate system architectures. This paper presents a method that helps simultaneously define the functional and design architectures of complex systems. Starting from an allocation matrix, or domain mapping matrix, which represents the couplings between the elements of two domains (functions vs. components), the method generates a sesign structure matrix (DSM) for each domain. A clustering algorithm is then applied to each DSM in order to generate a rearranged architecture. This method can be used in the synthesis process to provide architectural output that then forms the input data for the systems analysis process, where these data will help systems architects assess candidate architectures. We illustrate the method via an industrial case study focusing on a new automobile engine development project.
Acknowledgements
This work was carried out as part of a research project funded by a car manufacturer's powertrain architecture department. We would like to thank the senior managers and personnel of this department for their cooperation and assistance.