Hybrid feedforward-feedback active noise reduction for hearing protection and communication

Area of Science:

• Acoustical Engineering
• Signal Processing
• Biomedical Engineering

Background:

• Active noise reduction (ANR) systems are crucial for hearing protection in noisy environments.
• Traditional feedforward ANR systems face challenges with gain stability and low-frequency performance.
• Existing ANR architectures require improvements for effective noise cancellation in diverse conditions.

Purpose of the Study:

• To develop and validate a hybrid ANR architecture combining feedback and feedforward control.
• To enhance noise reduction performance, particularly at low frequencies, for hearing protection devices.
• To ensure the stability and efficiency of the ANR system during communication.

Main Methods:

• A hybrid ANR architecture integrating source-independent feedback ANR with a Lyapunov-tuned leaky LMS filter (LyLMS).
• Implementation of a filtered-X LyLMS for earplugs to manage nonconstant cancellation path gain.
• Utilizing a fast time-domain identification method for high-fidelity infinite impulse response (IIR) cancellation path modeling.

Main Results:

• The earcup achieved up to 40 dB noise reduction for specific tones and 30-32 dB for cockpit noise, with a 15 dB improvement at low frequencies.
• The earplug demonstrated 46-48 dB noise reduction for tones and 40-41 dB for helicopter noise, with minimal impact on speech communication.
• A significant improvement in the ANR system's stability metric, by factors of 2 to several orders of magnitude, was observed.

Conclusions:

• The hybrid ANR architecture offers superior noise reduction and stability compared to traditional methods.
• This technology provides effective hearing protection across a wide frequency range without compromising communication.
• The developed LyLMS filter and IIR modeling contribute to robust and efficient active noise control solutions.