On the Stability of Decentralized AVC/ASAC for Large-Scale Structures

Abstract

Active Vibration Control (AVC) and Active Structural Acoustic Control (ASAC) gained much attention in all kind of industries in the past. Promising results have been achieved in controlling the vibration and the noise emission/transmission of single panel structures. Especially for aircraft applications, concepts for the reduction of the turbulent boundary layer, rotor or jet noise are presented in the literature. In most cases the contributed work is focused on a single panel or a section of the fuselage/lining. However, an AVC/ASAC system can only be effective for the passengers when it is expanded to the entire fuselage structure. This expansion inevitably leads to a large number of sensors and actuators and thus to a controlled plant of high dimensions. For model-based control approaches especially, the system identification and the proof of stability would be challenging and probably not realizable. In this paper a strategy for such large-scale problems is investigated. A decentralized control approach with collocated actuator-sensor pairs is proposed. Since adjacent control loops are highly coupled by the underlying structure, special attention has to be given to the global stability of the entire control system. Instead of proving local stability and setting a global master gain, a method for the tuning of the single collocated control loops is developed that takes the cross-couplings into account. Based on data of DLR's experimental aircraft Dornier 728, it can be shown that the new method increases the performance of the control system compared to the master-gain method.