Active Noise Control Using Finite Element-based Virtual Sensors

Abstract

This paper describes results of an active noise control system using finite element-based virtual sensors. Systems for active reduction of noise are usually designed to minimize the noise at a physical error sensor. To obtain better noise reduction at a location away from the physical error sensors, virtual sensor techniques have been proposed. These methods require calibration with physical sensors at the monitor sensor location and virtual error sensor location. This can be time-consuming and inconvenient. This paper describes a method in which virtual sensors are determined using a finite element model to eliminate the calibration step. The method is tested in an experimental set-up for a ventilation system with heat recovery. Results are shown, comparing the performance of the finite element based virtual sensor to that of a physical sensor and a measured virtual sensor.