Head movement effects on the performance of an ANC headrest in a car cabin

Active noise control (ANC) headrest systems in car cabins use headrest loudspeakers as secondary sources to cancel noise, but the noise reduction performance deteriorates when the human head moves, and one of the most important reasons is the change in the secondary paths. The effect of the movement of the real human head on the performance of an ANC headrest in a car cabin is investigated in this paper and a database is presented for the first time. A total of 600 secondary paths from 2 headrest loudspeakers to 2 microphones fixed at the ears were measured with 10 subjects in 3 scenarios where the human head moved, and the ANC simulations based on the fixed-coefficient control strategy were carried out using these secondary paths and recorded road noise. The initial noise reduction is about 10.7 dBA, but it decreases by up to 0.7 dBA even when the subject tries to maintain the same position and posture, and by up to 2.5 dBA when the subject returns to the initial position and posture after some free body movements in the seat. For a subject who gets off the car, comes back in and tries to return to the initial position and posture, the deterioration in noise reduction can be up to 5.8 dBA. Measurements and simulations were also performed with a dummy head in the seat for comparison. The real-time ANC experiment results in an electric car were presented to support the findings in the simulations. This research indicates that robust adaptive algorithms with fast convergence speed are required in practical applications.