摘要患有阻塞性睡眠呼吸暂停综合症的病人通常都伴随有打鼾现象，用多导睡眠监测仪 对鼾症病人进行诊断是最常用的手段，但 PSG 价格昂贵且易造成患者的不适，而对鼾声 信号进行时域或频域的检测分析，相比于前者要更廉价且简便。鼾声信号作为语音信号 的一种，可以尝试提取其基音频率和共振峰，并加以分析，从基频和共振峰的角度寻找 鼾声信号特征与鼾症不同病因的关系。
本文依托国家自然科学基金面上项目（基于声学分析的鼾症人群鼾声来源及上气道 阻塞部位识别，No。61271410）,基于不同病因患者的鼾声数据，分析鼾声基音频率和共 振峰特征是否具有与不同病因机理的关联关系。首先阐述了提取基音频率和共振峰的几 种不同方法，详细介绍了本文用到的主要方法：LPC（线性预测编码）算法，并基于实测 的不同病因患者鼾声信号音频数据进行了特征提取及分析，初步分析结果表明，不同病 因患者鼾声信号的共振峰有一定区分度，基音频率暂未发现有明显区分度。考虑到当前 实测的患者鼾声数据样本量偏小，建议未来进一步增加样本量的同时开展新的特征集合 研究。75935
毕业论文关键词 阻塞性睡眠呼吸暂停/低通气综合症 基音频率 共振峰 线性预测编码
毕 业 设 计 说 明 书 外 文 摘 要
Title Analysis of snoring signals based on fundamental frequency and formant
Abstract Many patients with Obstructive Sleep Apnea/Hypopnea Syndrome have the symptoms of snoring。 PSG (Polysomnography)is commonly used for the clinical diagnosis of OSAHS。 However, PSG equipment is expensive and causes patients uncomfortable。 Meanwhile, analysis of snoring signals towards time domain and frequency domain is much cheaper and more operable 。Because snoring signals are speech signals, we can study them by analyzing data of their fundamental frequency and formant, by which means we may find the relationship between snoring signals and the different causes of OSAHS。
The work relying on the National Natural Science Foundation of China ， the recognition of snoring signal resource and upper airway obstruction site from snoring crowd based on the acoustic analysis ,No。61271410,studys the relationship between the fundamental frequency and formant and different casues of OSAHS based on different patients’ snoring data。 The paper elaborates some methods of obtaining fundamental frequency and formant, and elaborates Linear Prediction Coding in detail。 Preliminary studies indicates that there is some discrimination between different patients’ snoring but no obvious discrimination between fundamental frequency。 Sample size is too small to make a definitive conclusions。 Deeper study needs bigger sample size。
Keywords Obstructive Sleep Apnea/Hypopnea Syndrome Fundamental frequency Formant Linear Prediction Coding
本科毕业设计说明书 第 I 页
目 次
1 绪论 1
1。1 研究背景及意义 1
1。3 论文主要工作及结构安排 2
2 语音信号处理基础知识 4
2。1 基于基音频率和共振峰的鼾声特征分析:http://www.youerw.com/tongxin/lunwen_86999.html