Active control of low-frequency broadband air-conditioning duct noise

The low-frequency broadband noise generated by an air-conditioning unit on a railway vehicle may be significant in situations where the vehicle is stopped or its speed is low. Low-frequency acoustic radiation is extremely difficult to attenuate using passive means, so in this work, active noise control (ANC) techniques were applied. Laboratory experiments were performed in which noise from two different sources, namely from loudspeakers and an axial fan, was introduced into a real air-conditioning duct, identical to that installed on a subway vehicle. The control approach used a doubled, single-input, single-output (SISO), feedforward, filtered-X LMS algorithm. Control algorithms were designed using the Matlab-Simulink program and the Real Time Windows Target toolbox of Matlab in order to run the ANC in real time. Attenuations of 15-20 dB at the error microphone locations were achieved when loudspeakers were used as primary noise generators. However, noise reduction was quite poor when noise was introduced via the axial fan. This was due to turbulence generated by the airflow, which has a negative influence on the performance of the control system.