Systematic Identification of Decoupling in Dynamic System Models

Abstract

This paper proposes a technique to quantitatively and systematically search for decoupling among elements of a dynamic system model, and to partition models in which decoupling is found. The method can validate simplifying assumptions based on decoupling, determine when decoupling breaks down due to changes in system parameters or inputs, and indicate required model changes. A high-fidelity model is first generated using the bond graph formalism. The relative contributions of the terms of the generalized Kirchoff loop and node equations are computed by calculating and comparing a measure of their power flow. Negligible aggregate bond power at a constraint equation node indicates an unnecessary term, which is then removed from the model by replacing the associated bond by a modulated source of generalized effort or flow. If replacement of all low-power bonds creates separate bond graphs that are joined by modulating signals, then the model can be partitioned into driving and driven subsystems. The partitions are smaller than the original model, have lower-dimension design variable vectors, and can be simulated separately or in parallel. The partitioning algorithm can be employed alongside existing automated modeling techniques to facilitate efficient, accurate simulation-based design of dynamic systems.