Dynamic substructuring via the adaptive MCS algorithm

This paper describes some simple linear control strategies for dynamically substructured experiments. The purpose is to provide some benchmark strategies against which more complex substructured experimental systems may be judged. This increased complexity may be due to additional degrees-of-freedom in the substructured system, continuous variations in the dynamic parameters, unknown dynamics, external disturbances and the introduction of nonlinear elements into the system. Additional complexity may also be introduced by more sophisticated control policies. This paper introduces such a controller, which is an extension of the adaptive minimal control synthesis (MCS) algorithm, called the error-driven MCS with integral action (Er-MCSI) algorithm. Simulation results show that Er-MCSI significantly outperforms even a "perfectly" tuned and robust benchmark controller as system uncertainty increases.