ACTIVE NOISE CONTROL IN A CAVITY USING INTERPOLATED MODELS

This paper demonstrates a practical way for system identification using subspace and cubic spline methods to minimize the task of obtaining models using experimental frequency responses for a large number of sensor-actuator locations. These models are basically required in varied combinations for optimal actuator-sensor placement. Measuring at all the required positions is impractical and analytical modelling methods are helpless for nonideal boundary conditions. This paper also presents experimental results of active noise control, using interpolated models for an acoustic cavity. Acoustic enclosures are resonant systems which required an accurate system model to design high performance controllers. A controller using minimax LQG control method based on interpolated models provides promising results for the use of interpolated models in control design.