An improved algorithm for liner impedance eduction in low signal-to-noise ratio flow duct

Abstract

The impedance eduction technique is widely used by the aeroacoustics community to obtain liner property in a flow duct. However, the obtained impedance is often found to be discontinuous in the frequency domain which violates theoretical liner models. The low signal-to-noise ratio (SNR, in dB) is one factor leading to this unexpected result. To overcome this, a weighting coefficient, represented by an SNR dependent sigmoid function with two control parameters, is introduced to the cost function in the iteration process. The proposed algorithm is employed to measure the impedances of two liners and results show an improvement in the smoothness of the resultant impedance curves over those obtained from conventional cost function.