A low-cost, portable, active noise control unit : an optimal fixed-filter active noise control solution for construction noise sources

Abstract

Noise pollution affects millions of people daily and can also cause high blood pressure, lack of sleep, heart disease, and stress. In Singapore, one of the most common form of noise pollution is construction noise. Construction sites can have noise levels that is deemed hazardous and consist of transient situations where different trades come to the site to perform their work for at any point of time. The aim of this study is to investigate the most optimal fixed-filter Active Noise Control (ANC) system setting which can be used to improve the conventional fixed-filter ANC’s performance. This will involve data measurements on different noise sources used to train the fixed-filter ANC system and performance evaluation using an algorithm developed in MATLAB. The noise sources will also be tested with a real time fixed-filter ANC system with its performance data being collected as well. The collected data from both platforms will be used to perform feature extraction, classification, and validation. The performance indicators will then be reviewed are overall accuracy and reliability. After running through the MATLAB simulations, the results have shown that different noise sources have selective suitability to different construction noises. A total of 5 noise sources for training the ANC system and 13 construction noise samples were used in the analysis. Research findings are as reported in the conclusion section of this report.