Design and analysis of an improved hybrid active noise control system

Countering mutual coupling in the inherent structures and the non-availability of proper reference or error signals for the operation of adaptive filters has been a challenge in the design of a Hybrid active noise control (HANC) system. Existing HANC systems have not simultaneously addressed these two issues. In this paper, we propose an improved HANC system which generates most appropriate signals for operating the adaptive filters along with decoupling the inherent feedforward and feedback ANC systems in the HANC system. Here, we use an extra adaptive filter which post-processes the error signal to obtain true signals for operating the feedforward control filter which improves the convergence speed and the overall noise reduction. A Joint-Optimized Normalized Least Mean Squares algorithm is employed for generating the true signal by optimizing the system misalignment, which further improves the convergence of the adaptive filter in the feedforward part. The adaptive filter in the feedback structure also gets appropriate signals by internally generating a reference signal which is independent of the feedforward structure. A complete mathematical analysis of the proposed algorithm is done. Simulation results validate the advantages of the proposed system over its counterparts at the cost of increased computational complexity.