Comparative Study of Adaptive Algorithms for Vehicle Powertrain Noise Control

Active noise control systems have been gaining popularity in the last couple of decades, due to the deficiencies in passive noise abatement techniques. In the future, a novel combination of passive and active noise control techniques may be applied more widely, to better control the interior sound quality of vehicles. In order to maximize the effectiveness of this combined approach, smarter algorithms will be needed for active noise control systems. These algorithms will have to be computationally efficient, with high stability and convergence rates. This will be necessary in order to accurately predict and control the interior noise response of a vehicle. In this study, a critical review of the filtered-x least mean square (FXLMS) algorithm and several other newly proposed algorithms for the active control of vehicle powertrain noise, is performed. The analysis examines the salient features of each algorithm, and compares their system performance. Numerical simulations utilizing synthesized data, are conducted to study the convergence rates of these algorithms. These convergence rates are critical for the noise control outcome. Furthermore, measured powertrain noise response is employed, to verify the system's performance under more realistic conditions. The individual engine orders are targeted for attenuation or enhancement, to achieve the desired vehicle interior sound quality.