The method of identifying the dominative substructure for dynamically substructured systems

Dynamically substructured system (DSS) technique is increasing recognized in the testing field. A DSS method decomposes an entire system into several substructures, which are tested numerically or physically, but run as a whole like the original system. Thus, the advantages of DSS include flexibility, space-saving, and cost-reduction. In a DSS test, actuators produce unexpected disturbance, when transferring mechanical power and control signals between numerical and physical substructures. The ill-conditioned substructure that contains real-time disturbance degrades the computation and control accuracy, and in turn, may corrupt the whole test. This paper refers such substructure to “dominative substructure”, and proposes a method to identify it. The techniques of Lyapunov function, Riccati-like equation, and linear matrix inequalities are utilized to assess the substructure robustness. The robust indices can recognize the dominative substructure prior to practice, which monopolizes the overall DSS testing efficiency, and give a clue about how to effectively improve the DSS set up. Numerical results based on a multivariable mass-spring-damper DSS are presented to demonstrate the proposed method.