Abstract:

Virtual sensors using the forward-difference prediction algorithm have shown great promise for remote sensing in active noise control systems. These virtual sensors use linear arrays of microphones containing two or more microphone elements. Under certain circumstances the accuracy of these virtual sensors has been disappointing due to inherent errors such as phase and sensitivity mismatches as well as uncorrelated noise between the microphone elements. The open loop sensing microphone weights are determined without accounting for these errors, which can lead to inaccurate pressure estimates at the virtual location. It is hoped that an adaptive LMS algorithm will overcome these difficulties by using the algorithm to adapt the weights used to modify the signals from the sensing elements which form the array. The algorithm adapts the sensing microphone signal weights until the weighted sum of the sensing microphone signals produces the same signal as the microphone at the virtual location. Once this has been achieved,the sensing microphone weights are fixed and the microphone at the virtual location is removed, thus creating a virtual microphone. The quality of the pressure estimate and the level of control achieved at the virtual location using an open loop algorithm is compared with that obtained using the LMS adaptive algorithm.