论文部分内容阅读
通过对36名正常组和30名中、轻度病态嗓音组成年女性的频域相对信噪比进行分析,结果表明:正常嗓音的频域相对信噪比呈一个稳定的分布,而病态嗓音的相对信噪比在大于1700Hz频段与正常对照组有显著性差异(P<0.01).在此基础上,引入模糊数学的隶属函数(μ)来描述颇域相对信噪比的分布,此函数能够准确直观的定量表示噪声分布情况,可以认为(μ)函数可作为客观定量评价嗓音素质的有效参数.
The frequency-domain relative signal-to-noise ratio of 36 normal subjects and 30 moderate and mild morbid voices was analyzed. The results showed that the frequency-domain relative signal-to-noise ratio of normal voice showed a stable distribution, while the morbid voice The relative signal-to-noise ratio (SNR) was significantly different from that of the normal control (P <0.01) in the frequency band of more than 1700 Hz. Based on this, the membership function (μ) of fuzzy mathematics was introduced to describe the relative signal- Accurate and intuitive quantitative representation of the noise distribution, it can be considered (μ) function can be used as an objective and quantitative evaluation of effective voice quality parameters.