Using frequency-dependent causality analysis and automated tuning with broadband ANC systems to optimize the performance of the 3D sound field in a passenger vehicle

摘要

As broadband active noise control systems rapidly approach industrialization in the automotive sector, there is an emerging trend in system architectures that often result in limited performance. While past studies have focused on numerical analyses to assess theoretical performance limits, this study focuses on experimental methods to assess the general and frequency-dependent causality criteria in order to identify deficiencies in the system architecture that can hinder performance. Prediction methods and a nonlinear optimization algorithm are used in tandem to optimize the performance for the 3D sound field inside the vehicle based on customer-driven target objectives. This prediction/optimization method can identify new system architectures that satisfy the causality criteria, and/or identify new tuning parameters to work around the causality violations, thus ensuring optimal system performance.