Course program and timetable

The course presents the fundamental elements related to the active control of acoustic noise, with the final aim of providing students with the conceptual and methodological tools in the design of active control systems. For this purpose, adaptive digital filtering algorithms and disturbance rejection methods are introduced. Both broadband and narrowband active noise control methods are presented, using feedforward and feedback, single and multi-channel implementation schemes. The nonlinear active noise control problem is also briefly discussed, as well as the rejection of non-gaussian impulsive noise. Numerical and implementation-related issues are also reviewed.

The course is structured as 6 blocks of 4-hour lessons.

1. WED 14/01, 9.30-13.30. Acoustic noise models. Passive, semi-active and active control. Active control schemes: feedforward/feedback control, broadband/narrowband control, single/multiple channel control. Applications. Interference in plane sound fields. Limits of performance of active noise control methods.
2. THU 15/01, 9.30-13.30. Adaptive filtering. Standard and modified LMS algorithm. Convergence analysis. Broadband feedforward control algorithms: FXLMS and leaky FXLMS.
3. WED 21/01, 9.30-13.30. The effect of acoustic feedback. Filtered-U Recursive LMS. FxdLMS and FxdLMS/CE.
4. THU 22/01, 9.30-13.30. Narrowband feedforward control. Periodic disturbance rejection. The waveform synthesis method. Adaptive notch filtering. Adaptive Noise Equalizer. Disturbance frequency estimation (methods by Regalia and Bodson).
5. WED 28/01, 9.30-13.30. Feedback control methods. Multiple channel control. Secondary path modeling.
6. THU 29/01, 9.30-13.30. Numerical and implementation-related issues: finite precision effects, quantization, automatic gain control. The nonlinear active noise control problem. Active noise control of impulsive noise signals.

All lessons will take place in the Sala Seminari of the Dip. di Elettronica, Informazione e Bioingegneria of the Politecnico di Milano.

Lectures

0. Course presentation
1. Introduction
   
   • Introduction to active control of noise and vibrations
   • ANC applications
2. Basics of acoustics
3. Interference in plane waves
4. Adaptive filtering
5. Broadband feedforward ANC
   
   • Part 1: The FxLMS algorithm
   • Part 2: The FuLMS algorithm
6. Narrowband feedforward ANC
   
   • Narrowband feedforward ANC
   • Appendix - Adaptive notch filtering with frequency tracking
7. Multiple-channel ANC
8. Feedback ANC
9. On-line secondary path modeling techniques
10. Implementation issues
    
    • Implementation issues
    • Appendix - Direct digital implementation of notch filters
11. Active control of impulsive noise
12. Nonlinear active noise control

Software and teaching material

• Matlab m-files for single and multiple source simulation
• Matlab m-files for ANC examples
• Java applets for harmonic plane wave interference simulation, by Andrea Castelli and Filippo Viganò (the applets require Sun Java VM 5.0 or higher). A downloadable version is also available here.