Harmonic narrowband active noise control system with linear and nonlinear phase secondary path modeling for rotating machinery

Abstract

In light of the harmonic characteristics often exhibited by narrowband noise from rotating machinery, this paper proposes a harmonic narrowband active noise control (HNANC) system to reduce the computational burden caused by filtering the secondary path modeling. The innovation lies in the method of secondary path estimation filtering: when the secondary path has linear phase characteristics, only the fundamental narrowband component is filtered, regardless of the total number of narrowband components present. Conversely, when the secondary path exhibits nonlinear phase characteristics, only a selective subset of the narrowband components undergoes the secondary path estimation filtering process. This approach notably curtails computational complexity, particularly in environments featuring multiple frequency noise components. Furthermore, an online local secondary estimation method only for a selective subset of the narrowband components is introduced in the HNANC framework, which further reduces the computational complexity of the NANC with online secondary modeling. To substantiate its effectiveness, the convergence and steady-state error of the conventional method, its optimized version, and HNANC using both offline and online, as well as linear and nonlinear phase secondary path modeling, are compared across various narrowband noise environments. Moreover, the proposed HNANC system, both with offline and online secondary modeling, undergoes testing on a duct ANC platform for exhaust fan noise. This practical validation helps underscore the potential real-world applicability of the HNANC system.