论文部分内容阅读
针对宽带噪声的频谱特点和人耳听觉的掩蔽特性,提出了一种噪声掩蔽的等效原则,以及一种新的烦恼度指数的计算方法。在采用时频分析的方法,结合统计学规律,辨识出噪声中致人烦恼的频率成份之后,利用该烦恼度指数可度量各频率成份的烦恼程度,从而建立了一种噪声品质评估模型。试验研究的结果表明:基于掩蔽特性的噪声品质评估模型可以快捷而准确地识别噪声中烦恼的频率成份,并与人耳主观辨识的结果具有良好的一致性。
Aiming at the spectral characteristics of broadband noise and the masking properties of human hearing, a new equivalent principle of noise masking and a new calculation method of annoyance index are proposed. After using the method of time-frequency analysis and the statistical law to identify the frequency components causing annoyance in the noise, the annoyance index can be used to measure the degree of annoyance of each frequency component to establish a noise quality assessment model. The experimental results show that the noise quality assessment model based on masking characteristics can quickly and accurately identify the frequency components fretted in noise and have good consistency with the results of subjective human recognition.