Abstract
近來，機械噪音的改善逐漸受到重視。而要改善機器的噪音，必先知道其噪音源的位置及其所輻射聲功率的大小。以往採用的聲強法雖有其可靠性，但對於無法隔離的噪音源系統，仍然無法正確地辨識各噪音源的聲功率值。本論文的主要目的就是以音源模擬法來辨識一噪音源系統的各噪音源聲功率值。此法的最大優點，就是對於無法隔離各噪音源的噪音源系統，仍能辨識各噪音源的聲功率值。除此，尚可減少量測的時間，以改善聲強法的缺點。最後並將此法應用於噴墨印表機的噪音源辨識工作上，並探討此印表機的發聲機制及改善方法。Lately, many reports attach importance to reduce the noiseradiated by working machine. From the view point of noise re-duction, it is necessary to locate the source position andknow how much power does it radiated. The intensity method,although, has its reliability, for a noise system which can'tisolate any source, the method is failed to identify the sourcepower. In this work, a more convenient, time saving method wasinvestigated to identify every source power of a noise system.The most important feature of this method is without using anyisolation techniques, one can still identify every source powerunder its normal working condition. At last, we apply thismethod to identify the source power of a jet printer, discussthe mechanism by which the sound radiated and how to improve it.