During the last years, more and more mechanical applications saw the introduction of active control strategies. In
particular, the need of improving the performances and/or the system health is very often associated to vibration
suppression. This goal can be achieved considering both passive and active solutions. In this sense, many active control
strategies have been developed, such as the Independent Modal Space Control (IMSC) or the resonant controllers (PPF,
IRC, . . .). In all these cases, in order to tune and optimize the control strategy, the knowledge of the system dynamic
behaviour is very important and it can be achieved both considering a numerical model of the system or through an
experimental identification process. Anyway, dealing with non-linear or time-varying systems, a tool able to online
identify the system parameters becomes a key-point for the control logic synthesis.
The aim of the present work is the definition of a real-time technique, based on ARMAX models, that estimates the
system parameters starting from the measurements of piezoelectric sensors. These parameters are returned to the control
logic, that automatically adapts itself to the system dynamics. The problem is numerically investigated considering a
carbon-fiber plate model forced through a piezoelectric patch.