Neural Based Active Noise Controller

In conventional active noise control (ANC) system, the primary noise is attenuated over the frequency band of interest based simply on the error signal, which does not take into account the human perception of noise. Hence, researchers have developed psychoacoustic ANC system to improve its noise reduction performance from psychoacoustic point of view. However, in this psychoacoustic ANC system, there may be a disturbance that is uncorrelated with the primary noise at the error sensor, which can severely degrade the system's noise reduction performance. Hence, in this paper, a psychoacoustic hybrid ANC system is proposed, which can simultaneously control both the correlated primary noise and uncorrelated disturbance from psychoacoustic point of view. Loudness is used as the psychoacoustic criterion for evaluating the system's noise reduction performance. Simulation results show the effectiveness of the proposed psychoacoustic hybrid ANC system.

Multichannel active noise control (ANC) systems have been investigated to reduce outdoor low-frequency noise, such as power transformer noise and traffic noise. However, in outdoor environments, wind noise is inevitable and significantly degrades the noise reduction performance of active control systems. Despite broad adoption in windy environments, porous microphone windscreens cannot eliminate low-frequency wind noise in microphones, especially when the wind speed is high. Assuming wind noise is additive and uncorrelated with the primary noise to be controlled, this paper analyses how the wind noise detected by the reference microphone impacts the control filters and the residual noise of a multichannel ANC system based on the multichannel filtered-reference least mean squares algorithm (MFxLMS). By extending our recent research on wind noise suppression in single-channel active noise control systems [Y. Chu, et al., J. Acoust. Soc. Am. 152(6), 3340–3345 (2022)], a modified MFxLMS algorithm based on a new cost function is proposed. Simulations with real-recorded wind noise are performed to verify the theoretical analysis.

Multichannel feedforward active noise control system combined with noise source separation by microphone arrays

Based on the theoretical analysis of literature, saturation in measured signal of active noise control (ANC) systems will degrade the convergence speed. However, the experiments show that the saturated input signal can speed up the convergence of the narrow-band ANC systems. This paper intends to remodel the saturation effects for feedforward and feedback ANC systems. Combining the action of analog-to-digital converters (ADC), the mathematical expression and block diagrams are proposed to model the saturation effects in the practical ANC systems. The derivation and simulation results show that since the saturation is able to amplify the principle component of signal, the convergence would be speeded up.

Electronic delay has been a critical problem in active noise control (ANC) systems. This is true whether a feedforward structure or a feedback structure is adopted. In particular, excessive delays would create a causality problem in a feedforward ANC system of a finite-length duct. This paper suggests a multirate signal-processing approach for minimizing the electronic delay in the control loop. In this approach, digital controllers are required in decimation and interpolation of discrete-time signals. The computation efficiency is further enhanced by a polyphase method, where the phases of low-pass finite impulse response (FIR) filters must be carefully designed to avoid unnecessary delays. Frequency domain optimization procedures based on H1, H2, and Hinfinity norms, respectively, are utilized in the FIR filter design. The proposed method was implemented by using a floating-point digital signal processor. Experimental results showed that the multirate approach remains effective for suppressing a broadband (200-600 Hz) noise in a duct with a minimum upstream measurement microphone placement of 20 cm.

In practical active noise control (ANC) systems, the primary path and the secondary path may be nonlinear and time-varying. It has been reported that the linear techniques used to control such ANC systems exhibit degradation in performance. In addition, the actuators of an ANC system very often have nonminimum-phase response. A linear controller under such situations yields poor performance. A novel functional link artificial neural network (FLANN)-based simultaneous perturbation stochastic approximation (SPSA) algorithm, which functions as a nonlinear mode-free (MF) controller, is proposed in this paper. Computer simulations have been carried out to demonstrate that the proposed algorithm outperforms the standard filtered-x least mean square (FXLMS) algorithm, and performs better than the recently proposed filtered-s least mean square (FSLMS) algorithm when the secondary path is time-varying. This observation implies that the SPSA-based MF controller can eliminate the need of the modeling of the secondary path for the ANC system.

It is important to remove large environmental noise in measurement using SQUIDs without magnetic shielding. Here, the environmental noise has been reduced by an active noise control (ANC) system in the MHz radio-frequency region. The ANC system including a phase and amplitude control circuit has been designed to allow precise adjustments of the phase and the amplitude of the anti-phase wave against the noise waves in the MHz region. In this paper, a sinusoidal wave in the MHz range was used as the radio frequency interference. When a SQUID was used for a signal detector, the ANC system suppressed this sinusoidal wave to the white noise level by about 54 dB. The performance of the ANC system for the chemical detection by nuclear quadrupole resonance (NQR) was checked in a simulated environment. One was the case of a real AM radio broadcasting signal with the dummy NQR signal and another was the real NQR signal with a sinusoidal wave as noise. For both cases, the ANC showed good performance for MHz range noise cancellation.

Infant incubators are widely used in neonatal intensive care units (NICU) for the pre-term or sick babies. In this paper, we propose an active noise control (ANC) system integrated with infant cry detection and classification function for infant incubators. The incubator ANC system reduces the high noise level which results in numerous adverse health effects for infants, and the cry detector and analyzer monitors the infants' physical conditions using the same ANC system hardware. The infant crying signals picked up by the error microphone of ANC inside the incubator are detected, the signals' features are extracted by using linear prediction coding (LPC), and Mel-frequency cepstral coefficients (MFCC) methods. The database of different signals and their associated infant health features will be established through training phase. The signal classification algorithm is based on shortest distance and Bayesian classifier. Simulation is conducted to evaluate the performance of proposed system using a Giraffe incubator from GE Healthcare. The experiments show that the proposed ANC system can reduce the harmful noise and the cry classification algorithm can recognize infant cries in order to develop the innovative, low cost methods to monitor the infant's health conditions.

The intake system in an automotive engine has a short duct compared with that of the exhaust system. The filtered-x LMS (FX-LMS) algorithm has been applied to the active noise control (ANC) system in a short acoustic duct. This algorithm design is based on the FIR (finite impulse response) filter; however, it has a slow convergence issue due to a large number of zero coefficients. To improve the convergence performance, the step size of the LMS algorithm was modified from fixed to variable. However, this algorithm is still not suitable for the ANC system of a short acoustic duct because the reference signal is affected by the backward acoustic wave propagated from a secondary source. Therefore, the recursive filtered-u LMS algorithm (FU-LMS) based on the infinite impulse response (IIR) is developed to consider backward acoustic propagation. Generally, this algorithm has a stability problem. The stability issue was improved using an error-smoothing filter. In this paper, the recursive LMS algorithm with a variable step size and smoothing error filter is designed. This recursive LMS algorithm, the FU-VSSLMS algorithm, uses an IIR filter. With fast convergence and good stability, this algorithm is suitable for the ANC system in a short acoustic duct, such as the intake system of an automotive engine. This algorithm is applied to the ANC system of a short acoustic duct. The disturbance signals used as primary noise source are a sinusoidal signal embedded in white noise and the chirp signal, which has a variable instantaneous frequency. The test results demonstrate that the FU-VSSLMS algorithm has a superior convergence performance when compared with the FX-LMS and FX-LMS algorithms. The algorithm can be successfully applied to the ANC system in a short duct, such as the intake duct.

An important step for active noise control (ANC) systems to be practical is to develop model independent ANC (MIANC) systems that tolerate parameter variations in sound fields. Reliabilities and stabilities of many MIANC systems depend on results of online system identifications. Parameter errors due to system identifications may threaten closed-loop stabilities of MIANC systems. A self-learning active noise control (SLANC) system is proposed in this study to stabilize and optimize an ANC system in case identified parameters are unreliable. The proposed system uses an objective function to check closed-loop stability. If partial or full value of the objective function exceeds a conservatively preset threshold, a stability threat is detected and the SLANC system will stabilize and optimize the controller without using parameters of sound fields. If the reference signal is available, the SLANC system can be combined with a feedforward controller to generate both destructive interference and active damping in sound fields. The self-learning method is simple and stable for many feedback ANC systems to deal with a worst case discussed in this study.

The objective of this research project is to develop an Active Noise Control (ANC) system based on the Filtered-X Least Mean Square algorithm. The ANC system developed here will be installed in a rocket engine testing facility of SpaceX Company near McGregor, Texas. As the SpaceX Company’s rocket engines have been increased in their powers, the engine noise from the testing facility is increasing rapidly. Although there is no noise complain currently from the nearby communities in McGregor, concerns are expected in the near future that the noise could possibly disturb people in McGregor. The ANC system can reduce the noise level using a few microphones and speakers connected to a DSP frontend, while a Passive Noise Control system would be much larger and require a higher cost to be implemented. This ANC system is currently implemented as a Single Input and Single Output system in a laboratory, effectively reducing the noise generated from a loudspeaker driven by the recorded rocket engine noise signal. Then, the system will be improved and operated as a Multiple Inputs and Multiple Outputs system to successfully adapt various wind, temperature, and humidity changes, leading to the maximum noise reduction toward McGregor.

Affine projection algorithms (AP) are widely used to implement active noise control (ANC) systems, due to their fast convergence speed and robustness, nevertheless their high computational cost could restringe their use in real-time applications. For that reason several low cost AP algorithms have been proposed, such as the Error Coded Affine Projection (ECAP) algorithm; however, in order to implement the algorithm in an ANC system is necessary to perform some modifications. In this paper, we present a version of the ECAP algorithm for single-channel ANC systems. The proposal is done using the conventional filtered-x scheme; furthermore, we use an evolutionary method to reduce the projection order of the algorithm during the computation process. Simulation results show a similar convergence speed to the AP algorithm, a lower final residual error and significant computational savings.

This paper proposes a method for shortening the distance from a noise detection microphone to a loudspeaker forming the feedforward type of active noise control (ANC) system. In the ANC system, the primary noise detected by the noise detection microphone and the secondary noise provided by the loudspeaker is required simultaneously arriving at another microphone, called error microphone, monitoring the effect of the active noise control. However, the reverberation involved in the secondary path from the loudspeaker to the error microphone increase the secondary noise components arriving later than the primary noise. The late components are not only invalid for cancelling the primary noise but also impede the cancellation. This paper hence proposes a method reducing the components by forming the noise control filter synthesizing the secondary noise with the cascade connection of an infinite and a finite impulse response filters. The distance can be thereby shortened.

This paper presents the design of quiet-comfort headboard that uses an active noise control (ANC) system to reduce low-frequency environmental and snoring noises, thus enables the consumers to sleep in a serene environment. The ANC system uses adaptive feedback ANC algorithm with error microphones and loudspeakers mounted on the headboard. MATLAB simulations using the measured transfer functions from the experimental setup are used for designing adaptive algorithm. Real-time experiments using a floating-point digital signal processor demonstrate the designed headboard is effective for reducing snore and other undesired noises.

This paper presents an algorithm structure for an active noise control (ANC) system based on an improved equation error (EE) model that employs the offline secondary path modeling method. The noise of a compressor in a gas station is taken as an example to verify the performance of the proposed ANC system. The results show that the proposed ANC system improves the noise reduction performance and convergence speed compared with other typical ANC systems. In particular, it achieves 28 dBA noise attenuation at a frequency of about 250 Hz and a mean square error (MSE) of about −20 dB.