Active noise control using adaptive POLYnominal Gaussian WinOwed wavelet networks

An active noise control algorithm is introduced based on adaptive wavelet networks using rationale functions with second-order poles wavelets. A novel network structure is derived using a nonlinear static mapping cascaded with an infinite impulse response filter wherein only input at the current step is needed to generate the output of the next sample. Therefore, incorporating tap delayed line of input–output of the physical system can be eliminated, avoiding the use of multidimensional wavelet networks. Online dynamic back-propagation learning algorithms (based on gradient descent method) are applied to adjust the network parameters. The local convergence of the closed-loop system is proved using discrete Lyapunov function. Simulations are carried out to compare the performance of proposed methods with other nonlinear algorithms (e.g. FxBPNN, Volterra, FLNN, and RFFLNN). Experiments are then conducted to evaluate the developed algorithms. Both simulation and experimental results show the superior performance of proposed method in terms of fast convergence rate and noise attenuation while avoiding curse of dimensionality.

A survey on active noise control in the past decade–Part II: Nonlinear systems

A combined bilateral and binaural active noise control algorithm for closed-back headphones

Filtered-x generalized mixed norm (FXGMN) algorithm for active noise control

Active noise control (ANC) algorithms are generally designed to obtain the attenuation of gaussian noise signals using suitable adaptive signal processing algorithms. The rejection of non-gaussian impulsive noise signals represents a much more critical task, with respect to which standard ANC algorithms generally fail to provide a satisfactory solution, due to convergence and instability problems. This paper proposes a novel ANC algorithm for the attenuation of impulsive noise, based on the online estimation of an alpha-stable model of the noise probabilistic description. Simulation analysis shows that the presented approach provides some improvement with respect to competitor methods.

Commercial active noise and vibration control systems for vehicles are gradually developing, mainly in Japan, for both active noise control of interior sound and active control engine mounts [1]. There have been four Japanese commercial applications: • Active control for booming noise by Nissan in 1991, • Active vibration mount for idle noise and vibration by Toyota in 1997, • Active vibration mount for diesel engines by Nissan in 1998, • Active control for low-frequency road noise by Honda in 2000. Other manufacturers have, of course, developed similar systems, but these have not progressed to production applications. Additionally, active control engine mufflers using cancelling loudspeakers have been successfully demonstrated. A new type of active muffler, using a controlled oscillating butterfly valve in the exhaust line has been developed recently. The future of active control in vehicles is dependent on customers who are willing to pay the higher price for additional noise and vibration control, so that cost reductions are the essential factor in further commercial products. Trends towards fuel efficiency and lighter vehicles run counter to reducing noise and vibration. This has given new interest in active noise control and active control engine mounts as a solution to these conflicting issues.

Active Noise Control (ANC) systems employing adaptive filters suffer from stability issues in the presence of impulsive noise. Due to its simplicity and less computational complexity, the Filtered-x Least Mean Square (FxLMS) algorithm is the most widely used ANC algorithm which minimizes the mean square error of noise, but it lacks robustness and stability in presence of high impulses. To overcome this limitation, new methods must be investigated. A robust adaptive algorithm (Bhagyashri algorithm) for impulsive noise suppression is already proposed in literature. In this paper, Bhagyashri algorithm is tested in ANC domain and thus Filtered-x Bhagyashri (FxBhag) algorithm is proposed. Computer simulations are executed to verify the enhanced performance of the FxBhag algorithm. The statistics of impulsive noise is modeled by Symmetric α-stable (SαS) distributions. The suggested solution exhibits better stability and faster convergence with almost same computational complexity as that of standard FxLMS algorithm and its variants.

Convex combination of the FxAPV algorithm for active impulsive noise control

In this paper, we propose the constrained stability least mean square (CS-LMS) algorithm for multichannel active noise control (ANC). It is shown that the centralized version is able to have better attenuation and tracking performance for different step sizes, ideal and noisy plant models. It is also shown that the proposed distributed version using an incremental communication strategy can obtain a better noise reduction performance is a proper step is used. These performance improvements over the NLMS solutions are obtained with a price of a slight increase in the numerical complexity.

In this paper we consider active noise control (ANC) of impulsive noise having peaky distribution with heavy tail. Such impulsive noise can be modeled using non-Gaussian stable process for which second order moments do not exist. The most famous filtered-x least mean square (FxLMS) algorithm for ANC systems is based on second order moment of error signal, and hence, becomes unstable for the impulsive noise. Recently we have proposed variants of the FxLMS algorithm where improved performance has been realized either by thresholding the input data or efficiently normalizing the step-size for adaptation. In the practical ANC systems, these thresholding parameters need to be estimated offline and cannot be updated during online operation of ANC systems. Furthermore, normalizing the steps-size for an impulsive noise source would essentially freeze the adaptation for very large impulses. In order to solve these problems, in this paper we propose a novel approach for ANC of impulsive noise sources. The proposed approach is based on data-reusing (DR) type adaptive algorithm. The main idea is to improve the stability by normalizing the step-size, and improve the convergence speed by reusing the data. The computer simulations are carried out to verify the effectiveness of the proposed algorithm.

The paper proposes an algorithm for active noise control (ANC) using the least squares lattice (LSL) algorithm. As a main algorithm for ANC, the LMS algorithm has been used for its simplicity. However, the LMS algorithm has problems of both convergence speed and estimation accuracy in the case of the broadband noise such as road noise which is one of the passenger compartment noises in a car running on rough roads. In order to solve these problems, the LSL algorithm for ANC is considered. By computer simulations, the LSL algorithm turns out to be more effective than the LMS algorithm in both convergence speed and estimation accuracy, especially for the colored input signal used in ANC of road noise.

All-pass filtered x least mean square algorithm for narrowband active noise control

This student paper presents an understanding of Active Noise Control (ANC) systems especially focus on the physical geometry about the detector and the observer and develops a series of simulation experiments for the different type of signal to test the cancellation performance using feedforward active noise control procedure with the fixed controller proposed by Tokhi and Leitch in 1987. This study will provide a further demonstration for the control approach and a foundation for proposing new ANC algo-rithms in the future studies.

This paper reports on the development of “full-gradient” and “simplified-gradient” versions of the filtered-u algorithm for active noise control. After discussing the general principles of active noise control, the paper presents full mathematical derivations of the full-gradient, simplified-gradient, and Feintuch-based versions of the filtered-u algorithm, clearly showing the various levels of simplifying assumptions that are made along the way. Finally, some illustrative simulation results are presented.

Traditional active noise control (ANC) systems rely on adaptive filtering techniques that assume a linear relationship between signals. However, in practical implementations, nonlinear distortions, particularly those introduced by loudspeakers, can degrade control performance, limiting the effectiveness of conventional ANC approaches. This paper presents a hybrid ANC algorithm, GFANC-THFxNLMS, designed to enhance adaptability and robustness in complex acoustic environments, such as encapsulated structures with openings. The proposed method leverages deep learning to construct a dataset of pre-trained sub-control filters and employs a 1D convolutional neural network (CNN) to dynamically generate the most suitable filter for the incoming noise. An adaptive filtering stage then fine-tunes the control filter in real time, ensuring optimal noise attenuation. The performance of GFANC-THFxNLMS is evaluated against conventional ANC algorithms, namely GFANC and THFxNLMS, under both time-invariant and time-varying conditions. The results indicate that the proposed method achieves enhanced noise reduction, with 6.74 dB using THFxNLMS, 8.01 dB using GFANC, and 9.49 dB using GFANC-THFxNLMS, highlighting its effectiveness in noise mitigation.