Synthesis of a robust broadband duct ANC system using convex programming approach

Area of Science:

• Acoustics and Noise Control
• Control Systems Engineering
• Signal Processing

Background:

• Duct noise poses challenges in various industrial and environmental settings.
• Existing active noise control methods often struggle with broadband noise and system uncertainties.

Purpose of the Study:

• To develop a robust active controller for broadband noise attenuation in ducts.
• To ensure performance and stability despite plant uncertainties using H2 and H(infinity) control.

Main Methods:

• A spatially feedforward structure was employed for controller design.
• H2 cost function and H(infinity) constraint were integrated for robust stability.
• Controller synthesis involved convex programming via Q-parametrization and frequency discretization.
• Sequential quadratic programming was used to find the optimal controller.

Main Results:

• The controller was successfully implemented on a digital signal processor (DSP).
• Experimental validation confirmed the system's effectiveness.
• Maximal attenuation reached 16.5 dB, with a total attenuation of 5.9 dB in the 200-600 Hz band.

Conclusions:

• The proposed robust active controller effectively attenuates broadband noise in ducts.
• The H2/H(infinity) approach ensures reliable performance and stability in uncertain environments.
• Experimental results validate the practical applicability and efficiency of the developed system.