Robust interior noise control using system identification and linear matrix inequality (LMI) based control

Numerical investigations into robust noise control in a rotorcraft are presented based on system identification using experimental input-output data and linear matrix inequality (LMI) based control approach. A structural-acoustic system consisting of a rigid enclosure with one flexible wall is considered. Piezoceramic patches placed on the flexible wall of the enclosure are used as actuators and condenser microphones placed outside and inside the enclosure are used as sensors. The system identification is carried out using subspace-based deterministic stochastic realization algorithm and a reduced order model of the plant is identified in the frequency range 0-1000 Hz. Two control problems are formulated: 1) the disturbance is assumed to be available for feedback, and 2) the disturbance is assumed to be unknown. A controller based on the reduced order plant is designed using LMIs employing multiple objective H/sub /spl infin// control techniques. Numerical simulation results are presented and discussed.