Neuro-Fuzzy Network-Based Nonlinear Hybrid Active Noise Control Systems

Area of Science:

• Acoustics and Signal Processing
• Control Systems Engineering
• Artificial Intelligence in Engineering

Background:

• Active noise control (ANC) is crucial for mitigating environmental sound pollution.
• Existing ANC systems face challenges with robustness and manual fine-tuning in complex environments.
• Hybrid active noise control (HANC) integrates multiple strategies for improved performance.

Purpose of the Study:

• To develop a novel adaptive neuro-fuzzy network (ANFN) controller for enhanced hybrid active noise control (HANC).
• To improve the robustness and effectiveness of active noise suppression.
• To address nonlinearities and reduce manual adjustments in complex acoustic environments.

Main Methods:

• An adaptive neural network was designed to minimize the mean square error of residual noise.
• A fuzzy logic strategy was incorporated to handle environmental nonlinearities and reduce manual tuning.
• The stability of the proposed ANFN-based HANC controller was rigorously proven using Lyapunov theorem.

Main Results:

• Numerical simulations demonstrated the effectiveness of the proposed ANFN-based HANC method.
• The controller showed superior performance in active noise suppression compared to existing approaches.
• The system proved robust under various challenging noise signal conditions.

Conclusions:

• The proposed adaptive neuro-fuzzy network controller significantly enhances hybrid active noise control performance.
• This ANFN-based approach offers a robust and effective solution for real-world noise pollution suppression.
• The method successfully overcomes limitations of traditional ANC systems in complex environments.