Active noise control based on a low-cost embedded system platform

In today’s world, creating a low noise envoriment is of great interest for people’s working and living. In the traditional way, it is achieved by using passive noise reduction. For example, using specific materials for acoustic isolation and acoustic absorption. In order to achieve a good result with passive noise reduction. The cost may be very high. Especially in the low frequency noise isolation field. For example, the absorption coefficient of 50mm thick fiber-glass is 0.94 at noise frequency 4000Hz. However when the noise frequency is reduced to 125Hz, the absorption coefficient drops to 0.17. The absorption performance is decreased by five times. Now, there is an alternative solution to do the low frequency noise cancellation is using Active Noise Control (ANC) system. The basic theory of Active Noise Control is to capture noise source and then generate an anti-noise signal to interfere with the unwanted noise. The Active Nosie Control system performs better at low frequency noise control compared with passive noise control. The low frequency noise means that the corresponding sampling rate will be lower, which results in the lower computational load. In 1960s, LMS (Least mean square) algorithm was introduced to the ANC system. However, the LMS algorithm cannot solve the instability problem of the path between control source signal and error microphones. The next big achievement is the introduction of feedforward FxLMS, which was proposed by Morgan in 1980. In this project, the feedforward FxLMS is first tested in the simulations and then implemented in a real time duct experiment. The simulations are both conducted in MATLAB and C code with generated signals. Then the C code is implemented in Arduino Due platform to do the real time single tone cancellation in a duct. The cancellation performance will be measured by a SPL meter.Bachelor of Engineerin