| Summary: | An optimal filter design method for a virtual mechanical impedance (VMI) control system is presented in this report. Existing passive methods for reducing machinery noise have limited applications because of weight reduction, cooling, and appearance problems. The VMI control method, which does not require a sensor and can be implemented using a small number of actuators, is an active approach that can be used to replace passive methods when mechanical tuning is difficult. However, VMI control methods can only reduce single-tone noise. In this regard, we improved the VMI control system to reduce the radiated sound power of the enclosure panel at multiple frequencies. To achieve this, we propose an optimal filter design method for a multichannel feedback controller. Unlike conventional VMI control, the presented technique is capable of multifrequency control and considers the radiation efficiency, damping, and estimates the errors of the structural-acoustic system. The controller designed using the proposed method satisfies the required robustness and minimum steady-state error in the target frequency band. Thus, this filter design method facilitates the quantitative determination of several criteria that are needed to evaluate system parameter changes such as radiation efficiency, damping, and estimation error, which are overlooked in a standard active structural acoustic controller. The results obtained using a finite element-based model and experimental investigations show that the radiated sound power of the enclosure panel is reduced at multiple target frequencies after the implementation of control.
|
|---|
