A multi-harmonic amplitude and relative-phase controller for active sound quality control

• An adaptive algorithm is proposed for controlling amplitude and relative-phase.
• A number of harmonic components can be dealt with without interfering with each other.
• The proposed controller incorporates a strategy for facing impulsive disturbances.
• Reductions of up to 39 dB and relative-phase shifts between [−π,π][−π,π] are attained.
• SQ targets based on Zwicker-Loudness and Roughness can be attained.

Current active sound quality control systems aim at dealing with the amplitude level of the primary disturbance, e.g.   sound pressure, forces, velocities and/or accelerations, which implicitly leads to Loudness control, regardless of the spectral structure of the disturbance. As far as multi-harmonic disturbances are concerned, auditory Roughness, arguably the most related psychoacoustic metric with rumbling perception in passenger cars, can be tackled not merely by dealing with magnitudes but also with the relative-phase of the narrowband components. This paper presents an adaptive control scheme conceived for dealing with multi-harmonic disturbances, which features the independent amplitude and/or relative-phase control of the input periodic components and an improved robustness to impulsive events. The adaptive control scheme is based on a frequency-domain delayless implementation of the complex-domain, least mean squares algorithm, whereof its convergence process is improved by using a forgetting factor. The control capabilities are evaluated numerically for single- and multiple-harmonic disturbances, including realistic internal combustion engine sound contaminated with noise and by impulsive events. By using long transfer paths obtained from a real vehicle mock-up, sound pressure level reductions of 39 dBSPL and the ability to displacing the relative-phase of a number of narrowband components between [−π,π][−π,π] are accomplished by the proposed control scheme. The assessment of the results by using Zwicker-Loudness and auditory Roughness models shows that the proposed adaptive algorithm is able to accomplish and stably preserve various sound quality targets, after completion of the robust convergence procedure, regardless of impulsive events that can occur during the system operation.

Figure optionsDownload as PowerPoint slide