搜索

x

留言板

尊敬的读者、作者、审稿人, 关于本刊的投稿、审稿、编辑和出版的任何问题, 您可以本页添加留言。我们将尽快给您答复。谢谢您的支持!

姓名
邮箱
手机号码
标题
留言内容
验证码

语音信号序列的Volterra预测模型

张玉梅 胡小俊 吴晓军 白树林 路纲

引用本文:
Citation:

语音信号序列的Volterra预测模型

张玉梅, 胡小俊, 吴晓军, 白树林, 路纲

Volterra prediction model for speech signal series

Zhang Yu-Mei, Hu Xiao-Jun, Wu Xiao-Jun, Bai Shu-Lin, Lu Gang
PDF
导出引用
  • 对给定的英语音素、单词和语句进行了采集并完成预处理. 分别应用互信息法和Cao 氏法确定了实际采集的语音信号序列的延迟时间和嵌入维数, 以完成语音序列的相空间重构. 通过计算实际采集的语音信号序列的最大Lyapunov指数, 完成了语音信号的混沌特性识别, 判定其具有混沌特性. 引入Volterra级数, 提出了一种具有显式结构的语音信号非线性预测模型. 为克服最小均方误差算法在Volterra模型系数更新时固有的缺点, 在最小二乘法基础上, 应用基于后验误差假设的可变收敛因子技术, 构建了一种基于Davidon-Fletcher-Powell算法的二阶Volterra 模型(DFPSOVF), 并将其应用于具有混沌特性的语音信号序列预测. 仿真结果表明: DFPSOVF非线性预测模型对于单帧和多帧语音信号均具有更好的预测精度, 优于线性预测模型, 并且能够很好地反映语音序列变化的趋势和规律, 完全可以满足语音预测的要求; 可以根据语音信号序列的嵌入维数选取预测模型的记忆长度. 所提出模型可以为语音信号重构和压缩编码开辟一条新途径, 以改善语音信号处理方法的复杂度和处理效果.
    The given English phonemes, words and sentences are sampled and preprocessed. For these real measured speech signal series, time delay and embedding dimension are determined by using mutual information method and Cao's method, respectively, so as to perform phase space reconstruction of the speech signal series. By using small data set method, the largest Lyapunov exponent of the speech signal series is calculated and the fact that its value is greater than zero presents chaotic characteristics of the speech signal series. This, in fact, performs the chaotic characteristic identification of the speech signal series. By introducing second-order Volterra series, in this paper we put forward a type of nonlinear prediction model with an explicit structure. To overcome some intrinsic shortcomings caused by improper parameter selection when using the least mean square (LMS) algorithm to update Volterra model efficiency, by using a variable convergence factor technology based on a posteriori error assumption on the basis of LMS algorithm, a novel Davidon-Fletcher-Powell-based second of Volterra filter (DFPSOVF) is constructed and is performed to predict speech signal series of the given English phonemes, words and sentences with chaotic characteristics. Simulation results under MATLAB 7.0 environment show that the proposed nonlinear model DFPSOVF can guarantee its stability and convergence and there are no divergence problems in using LMS algorithm; for single-frame and multi-frame of the measured speech signals, when root mean square error (RMSE) is used as an evaluation criterion the prediction accuracy of the proposed nonlinear prediction model DFPSOVF in this paper is better than that of the linear prediction (LP) that is traditionally employed. The primary results of single-frame and multi-frame predictions are given. So, the proposed DFPSOVF model can substitute linear prediction model on certain conditions. Meanwhile, it can better reflect trends and regularity of the speech signal series and fully meet requirements for speech signal prediction. The memory length of the proposed prediction model may be selected by the embedding dimension of the speech signal series. The proposed model can present a nonlinear analysis and more valuable model structure for speech signal series, and opens up a new way to speech signal reconstruction and compression coding so as to improve complexity and process effect of speech signal processing method.
    • 基金项目: 国家自然科学基金 (批准号: 11502133, 11172342, 11372167, 61202153)、陕西省重点科技创新团队项目(批准号: 2014KTC-18)、西安市科技计划(批准号: CXY1437(1))和榆林市科技计划 (批准号: 2014cxy-09, sf13-43, 2012 cxy3-6) 资助的课题.
    • Funds: Project supported by the National Natural Science Foundation of China (Grant Nos. 11502133, 11172342, 11372167, 61202153), the Key Science and Technology Innovation Team in Shaanxi Province, China (Grant No. 2014KTC-18), the Science and Technology Plan of Xi'an City, China (Grant No. CXY1437(1)), and the Science and Technology Plan of Yulin City, China (Grant Nos. 2014cxy-09, sf13-43, 2012 cxy3-6).
    [1]

    Maragos P 2013 Appl. Soft Comput. 13 3314

    [2]

    Wu X J, Yang Z Z 2013 Appl. Soft Comput. 13 3314

    [3]

    Max A L 2011 Advances in Nonlinear Speech Processing 7015 9

    [4]

    Cheng X F, Zhang Z 2013 Acta Phys. Sin. 62 168701 (in Chinese) [成谢锋, 张正 2013 物理学报 62 168701]

    [5]

    Chen D Y, Liu Y, Ma X Y 2012 Acta Phys. Sin. 61 100501 (in Chinese) [陈帝伊, 柳烨, 马孝义 2012 物理学报 61 100501]

    [6]

    Iasonas K, Petros M 2005 IEEE Trans. Speech Audio Process. 13 1098

    [7]

    Sun J F, Zheng N H, Wang X L 2007 Singal Process. 87 2431

    [8]

    Maciej O 2005 Chin. Phys. 14 2181

    [9]

    Xiao X C, Li H C, Zhang J S 2005 Chin. Phys. 14 2181

    [10]

    Zhang J S, Li H C, Xiao X C 2005 Chin. Phys. 14 49

    [11]

    Thyssen J, Nielsen H, Hansen S D 1994 ICASSP 185

    [12]

    Sigrist Z, Grivel E, Alcoverro B 2012 Signal Process. 92 1010

    [13]

    Mathews V J 1991 IEEE Signal Process. Mag. 8 10

    [14]

    Wei R X, Han C Z, Zhang Z L 2005 Acta Electron. Sin. 33 656 (in Chinese) [魏瑞轩, 韩崇昭, 张宗麟 2005电子学报 33 656]

    [15]

    Guerin A, Faucon G, Le Bouquin-Jeannes R 2003 IEEE Trans. Speech Audio Proc. 11 672

    [16]

    Zhang Y M, Wu X J, Bai S L 2013 Acta Phys. Sin. 62 190509 (in Chinese) [张玉梅, 吴晓军, 白树林 2013 物理学报 62 190509]

    [17]

    Henry D, Abarbanel N M, Rabinovich M I, Evren T 2001 Phys. Lett. A 281 368

    [18]

    Cao L Y 1997 Physica D 110 43

    [19]

    Rosenstein M T, Collins J J, de Iuca C J 1993 Physica D 65 117

    [20]

    de Campos M L R, Antoniou A 1997 IEEE Trans. Circ. Syst. 44 924

  • [1]

    Maragos P 2013 Appl. Soft Comput. 13 3314

    [2]

    Wu X J, Yang Z Z 2013 Appl. Soft Comput. 13 3314

    [3]

    Max A L 2011 Advances in Nonlinear Speech Processing 7015 9

    [4]

    Cheng X F, Zhang Z 2013 Acta Phys. Sin. 62 168701 (in Chinese) [成谢锋, 张正 2013 物理学报 62 168701]

    [5]

    Chen D Y, Liu Y, Ma X Y 2012 Acta Phys. Sin. 61 100501 (in Chinese) [陈帝伊, 柳烨, 马孝义 2012 物理学报 61 100501]

    [6]

    Iasonas K, Petros M 2005 IEEE Trans. Speech Audio Process. 13 1098

    [7]

    Sun J F, Zheng N H, Wang X L 2007 Singal Process. 87 2431

    [8]

    Maciej O 2005 Chin. Phys. 14 2181

    [9]

    Xiao X C, Li H C, Zhang J S 2005 Chin. Phys. 14 2181

    [10]

    Zhang J S, Li H C, Xiao X C 2005 Chin. Phys. 14 49

    [11]

    Thyssen J, Nielsen H, Hansen S D 1994 ICASSP 185

    [12]

    Sigrist Z, Grivel E, Alcoverro B 2012 Signal Process. 92 1010

    [13]

    Mathews V J 1991 IEEE Signal Process. Mag. 8 10

    [14]

    Wei R X, Han C Z, Zhang Z L 2005 Acta Electron. Sin. 33 656 (in Chinese) [魏瑞轩, 韩崇昭, 张宗麟 2005电子学报 33 656]

    [15]

    Guerin A, Faucon G, Le Bouquin-Jeannes R 2003 IEEE Trans. Speech Audio Proc. 11 672

    [16]

    Zhang Y M, Wu X J, Bai S L 2013 Acta Phys. Sin. 62 190509 (in Chinese) [张玉梅, 吴晓军, 白树林 2013 物理学报 62 190509]

    [17]

    Henry D, Abarbanel N M, Rabinovich M I, Evren T 2001 Phys. Lett. A 281 368

    [18]

    Cao L Y 1997 Physica D 110 43

    [19]

    Rosenstein M T, Collins J J, de Iuca C J 1993 Physica D 65 117

    [20]

    de Campos M L R, Antoniou A 1997 IEEE Trans. Circ. Syst. 44 924

  • [1] 王梦蛟, 周泽权, 李志军, 曾以成. 混沌信号自适应协同滤波去噪. 物理学报, 2018, 67(6): 060501. doi: 10.7498/aps.67.20172470
    [2] 田中大, 李树江, 王艳红, 高宪文. 短期风速时间序列混沌特性分析及预测. 物理学报, 2015, 64(3): 030506. doi: 10.7498/aps.64.030506
    [3] 王梦蛟, 吴中堂, 冯久超. 一种参数优化的混沌信号自适应去噪算法. 物理学报, 2015, 64(4): 040503. doi: 10.7498/aps.64.040503
    [4] 张玉梅, 吴晓军, 白树林. 交通流量序列混沌特性分析及DFPSOVF预测模型. 物理学报, 2013, 62(19): 190509. doi: 10.7498/aps.62.190509
    [5] 韩敏, 许美玲. 一种基于误差补偿的多元混沌时间序列混合预测模型. 物理学报, 2013, 62(12): 120510. doi: 10.7498/aps.62.120510
    [6] 高仕龙, 钟苏川, 韦鹍, 马洪. 基于混沌和随机共振的微弱信号检测. 物理学报, 2012, 61(18): 180501. doi: 10.7498/aps.61.180501
    [7] 张学清, 梁军. 风电功率时间序列混沌特性分析及预测模型研究 . 物理学报, 2012, 61(19): 190507. doi: 10.7498/aps.61.190507
    [8] 王国光, 王丹, 何丽桥. 混沌中信号的投影滤波. 物理学报, 2010, 59(5): 3049-3056. doi: 10.7498/aps.59.3049
    [9] 杜杰, 曹一家, 刘志坚, 徐立中, 江全元, 郭创新, 陆金桂. 混沌时间序列的局域高阶Volterra滤波器多步预测模型. 物理学报, 2009, 58(9): 5997-6005. doi: 10.7498/aps.58.5997
    [10] 陈 争, 曾以成, 付志坚. 混沌背景中信号参数估计的新方法. 物理学报, 2008, 57(1): 46-50. doi: 10.7498/aps.57.46
    [11] 王永生, 孙 瑾, 王昌金, 范洪达. 变参数混沌时间序列的神经网络预测研究. 物理学报, 2008, 57(10): 6120-6131. doi: 10.7498/aps.57.6120
    [12] 晋建秀, 丘水生, 谢丽英, 冯明库. 一种基于周期轨道统计的混沌信号不可预测性强弱的检测方法. 物理学报, 2008, 57(5): 2743-2749. doi: 10.7498/aps.57.2743
    [13] 李雪霞, 冯久超. 一种混沌信号的盲分离方法. 物理学报, 2007, 56(2): 701-706. doi: 10.7498/aps.56.701
    [14] 和红杰, 张家树. 基于混沌的自嵌入安全水印算法. 物理学报, 2007, 56(6): 3092-3100. doi: 10.7498/aps.56.3092
    [15] 闫 华, 魏 平, 肖先赐. 基于Bernstein多项式的自适应混沌时间序列预测算法. 物理学报, 2007, 56(9): 5111-5118. doi: 10.7498/aps.56.5111
    [16] 唐国宁, 罗晓曙. 混沌系统的预测反馈控制. 物理学报, 2004, 53(1): 15-20. doi: 10.7498/aps.53.15
    [17] 张家树, 李恒超, 肖先赐. 连续混沌信号的离散余弦变换域二次实时滤波预测. 物理学报, 2004, 53(3): 710-716. doi: 10.7498/aps.53.710
    [18] 汪芙平, 王赞基, 郭静波. 混沌背景下信号的盲分离. 物理学报, 2002, 51(3): 474-481. doi: 10.7498/aps.51.474
    [19] 韦保林, 罗晓曙, 汪秉宏, 全宏俊, 郭维, 傅金阶. 一种基于三阶Volterra滤波器的混沌时间序列自适应预测方法. 物理学报, 2002, 51(10): 2205-2210. doi: 10.7498/aps.51.2205
    [20] 张家树, 肖先赐. 用于混沌时间序列自适应预测的一种少参数二阶Volterra滤波器. 物理学报, 2001, 50(7): 1248-1254. doi: 10.7498/aps.50.1248
计量
  • 文章访问数:  3260
  • PDF下载量:  210
  • 被引次数: 0
出版历程
  • 收稿日期:  2014-11-12
  • 修回日期:  2015-06-02
  • 刊出日期:  2015-10-05

语音信号序列的Volterra预测模型

  • 1. 陕西师范大学, 现代教学技术教育部重点实验室, 西安 710062;
  • 2. 陕西师范大学计算机科学学院, 西安 710062;
  • 3. 西北工业大学自动化学院, 西安 710072;
  • 4. 西北工业大学电子信息学院, 西安 710072
    基金项目: 国家自然科学基金 (批准号: 11502133, 11172342, 11372167, 61202153)、陕西省重点科技创新团队项目(批准号: 2014KTC-18)、西安市科技计划(批准号: CXY1437(1))和榆林市科技计划 (批准号: 2014cxy-09, sf13-43, 2012 cxy3-6) 资助的课题.

摘要: 对给定的英语音素、单词和语句进行了采集并完成预处理. 分别应用互信息法和Cao 氏法确定了实际采集的语音信号序列的延迟时间和嵌入维数, 以完成语音序列的相空间重构. 通过计算实际采集的语音信号序列的最大Lyapunov指数, 完成了语音信号的混沌特性识别, 判定其具有混沌特性. 引入Volterra级数, 提出了一种具有显式结构的语音信号非线性预测模型. 为克服最小均方误差算法在Volterra模型系数更新时固有的缺点, 在最小二乘法基础上, 应用基于后验误差假设的可变收敛因子技术, 构建了一种基于Davidon-Fletcher-Powell算法的二阶Volterra 模型(DFPSOVF), 并将其应用于具有混沌特性的语音信号序列预测. 仿真结果表明: DFPSOVF非线性预测模型对于单帧和多帧语音信号均具有更好的预测精度, 优于线性预测模型, 并且能够很好地反映语音序列变化的趋势和规律, 完全可以满足语音预测的要求; 可以根据语音信号序列的嵌入维数选取预测模型的记忆长度. 所提出模型可以为语音信号重构和压缩编码开辟一条新途径, 以改善语音信号处理方法的复杂度和处理效果.

English Abstract

参考文献 (20)

目录

    /

    返回文章
    返回