Sequencing of information in modular model-based systems design

Abstract

Multi-disciplinary design of complex systems is characterized by many inter-dependencies between components. Therefore, adjusting their properties appropriately to satisfy all system requirements is difficult. Simulation models enable fast quantitative assessment of the system behavior. Unfortunately, monolithic system models are often not available, due to domain-specific knowledge that needs to be incorporated. This paper proposes a framework that enables a modular formulation of separate models that can be integrated easily into a full system model. Key ingredient is a model for dependencies between system and component attributes as polyhierarchies without feedback loops. The complex dependency structure of a system can be modularized and simplified. Modular models are integrated into a system model by sequencing the flow of information of the design task. The resulting system model has no circular dependencies and two distinct interfaces: independent design variables and dependent system performance measures. The approach is applied to two gear design tasks.