Noise control has generated increasing interest in recent years with increased concerns for environmental protection and industrial safety. The traditional method of noise reduction using absorbing materials and erected barriers works effectively only at middle and high frequencies. Thus, active method which is based on the principle of destructive interference between the primary noise source and the actively generated acoustic waves is introduced. This approach works best at low frequencies when the wavelength of noise is long compared to the dimensions of its surroundings....[ Read more ]

Noise control has generated increasing interest in recent years with increased concerns for environmental protection and industrial safety. The traditional method of noise reduction using absorbing materials and erected barriers works effectively only at middle and high frequencies. Thus, active method which is based on the principle of destructive interference between the primary noise source and the actively generated acoustic waves is introduced. This approach works best at low frequencies when the wavelength of noise is long compared to the dimensions of its surroundings.

In many existing active control methods, noise is reduced in very limited locations or along a single direction. For example, in active noise control inside passenger cars, the system is designed to reduce road noise level at passengers' seat locations; in noise reduction inside a duct, noise is reduced along the path of the duct. However, in most applications, it is desirable to reduce noise in a specific 3-D region. In view of this, a new model-based adaptive noise control system using loudspeaker array and error sensor array is developed in this thesis. First, open loop system transfer functions are designed using a theoretical propagation model. The transfer functions thus found are regarded as the nominal values for the complete system. Second, to compensate for deviations from the theoretical model, the transfer functions are adapted using error measures from error sensor array by LMS algorithm. In addition, a linear predictor is introduced into the system for feedback control, thus input signal driving the secondary sources is obtained from the error signals.

In this study, we concentrated on reduction of broadband noise with a maximum frequency of 1000 Hz. Theoretical and simulation results show that our approach is effective for noise reduction in three dimensional space. Experiments using real-time active noise control system confirms the performance of the designed system.